U.S. Army Medical Department, Office of Medical History
Skip Navigation, go to content







AMEDD MEDAL OF HONOR RECIPIENTS External Link, Opens in New Window






Chapter XXI





A brief outline of the organization, location, and scope of work of the experimental gas field, known as Hanlon Field, is given in Chapter II. Of the sections of the field, only the medical sections (pathological and physiological sections) come within the scope of the present volume.

Personnel detailed from the Medical Corps had been attached to the staff of the field, and had acted in an advisory capacity from a very early date. Yet, because of the technical nature of the equipment required and the consequent difficulty in obtaining it, the medical sections were practically the last to come into active operation at the field. Nevertheless, very detailed and extensive work was under way and completed in some parts at the time of the armistice.

This work had been separated and designated under the heads of the pathological section and physiological section. However, the total quantity of material collected to the close of the field was not of such volume as to make necessary a separate summary report under each section. Accordingly, the whole is considered here under one head, no effort being made to indicate the particular work done under the direction of the personnel of each section. In fact, the actual experiments were closely interrelated and were made with such cooperation that it would have been difficult to make such a separation in any case.

A very considerable part of the work which these sections were able to do was concerned with the study of the lesions produced by various vesicants, or of internal toxic effects produced by "gases" taken into the body of animals or man either through the respiratory organs or by intravenous injection. A number of these investigations were carried out by these sections solely to determine the pathological effects produced by various war gases for the sake of the data alone. Such studies were embodied in the reports of these sections
(a) This chapter comprises the full text of a special Summary Report of the Work of Hanlon Field (Experimental Gas Field), Chemical Warfare Service, American Expeditionary Forces, France, pathological and physiological sections, with certain changes in the arrangement.
The following special reports which were submitted, through military channels, to the Chief, Chemical Warfare Service, are included in this chapter:
Special Report No. 39: Effects of Intravenous Injections of Dichlorethylsulphide in Rabbits, with Special Reference to the Blood and Hematopoietic Tissues. Submitted by Maj. A. M. Pappenheimer. M. C., and Capt. Morgan Vance, M. C., November 10, 1918.
Special Report No. 38: The Cutaneous Lesions Caused by Dichlorethylsulphide and Lewisite in the Horse. Submitted by Maj. A. M. Pappenheimer, M. C., November 2, 1918.
Special Report No. 57: Toxicity of Dimethyltrithiocarbonate. Submitted by Maj. A. M. Pappenheimer, M. C., for Mal. H. C. Clark, M. C. The work was performed by Capt. S. Goldschmidt, M. C.; Capt. B. M. Vance, M. C.; and Capt. D. Ah. Wilson, M. C., November 29,1918.
Special Report No. 15: The Behavior of Certain Slugs and Snails in the Presence of Dichlorethylsulphide (HS) Submitted by Maj. H. C. Clark, M. C., in charge of the pathological section, August 20,1918.
Special Report No. 44: A Suit Designed by Capt. H. G. Knowland, C. W. S., and Second Lieut. T. M. Knowland, C. W. S., for protection against Mustard Gas. Submitted by Maj. A. N. Richards, S. C.; Capt. S. Goldschmidt, C. W. S.; and Capt. H. G. Knowland, C. W. S.; November 18, 1918.
Special Report No. 49: The Protective Power of Sag Paste, Calcium Hypochlorite Ointment, and Petrolatum against Dichlorethylsulphide (HS) and Lewisite I. Submitted by Maj. C. B. Kerr, M. C., November 26, 1918.


and covered specifically three compounds, namely, dichlorethylsulphide or mustard gas, Lewisite I, and dimethyltrithiocarbonate. In addition, a very large number of examinations of lesions and general body conditions were made as autopsies in the cases of animals "gassed" in course of general field experiments carried out in cooperation with other sections. Such examinations became of the nature of routine for these sections, and results from them were given in general in the reports covering the main experiments, hence issued by other sections, in particular, the field gas experimentation section.

The second phase of the work done by the medical sections was the study of various methods of gas defense through the medium of physiological tests. No work on masks, however, was included under this function. On the other hand, because of the great importance which mustard gas had acquired in operations, these studies were concerned almost entirely with defense against this vesicant. These investigations took up, first, certain methods of the detection or estimation of the strength of air concentration of mustard gas. The proposal, originating at the American University, to use the behavior of snails for the detection of mustard gas was considered. Other low-form organisms were also tested for the same characteristics. At the same time, an attempt was made to standardize for rapid estimations the relation between the concentration of mustard gas in the air mixture and degree of lesion produced on human skin. In addition to the work on the detection of mustard gas, several investigations were carried out on special equipment for protection against this gas. These comprised tests of treated fabrics, various skin ointments, of which "sag paste" is the best example, and "tar" materials for the protection of horses' feet. Finally, in the study of mustard gas defense the general problem of the treatment of lesions produced by the same gas had been taken in hand. As is well known, no effective solution of this important problem was ever reached, either by our own service, or that of our Allies or of the enemy.

A third phase of the activities of these sections was the observation of actual gas casualty cases in hospitals, and the collection and examination of specimens from autopsies. b This work was taken in hand quite late and never fully developed. This function of the medical services in the field would certainly have been extended had the war continued.



Two studies were made directly for determining the effects produced by dichlorethylsulphide, each concerning a rather special and hitherto untouched phase of the general question of mustard gas pathology.

The first of these was a short investigation to determine the accuracy of the conclusion arrived at in some experiments that the serum from mustard gas blisters was capable itself of a vesicant action. Twenty human subjects were blistered and the fluid applied to other areas on their own skin, as well as to the skin of fresh subjects. The results were entirely negative, and it was concluded that the serum from such blisters had no vesicant action.

The other investigation of this class was a full study of the effects of intravenous injections of dichlorethylsulphide in rabbits, with special reference to

b Certain autopsy protocols from the laboratory of Hanlon Field are included in Chapter VI.


effects on the blood and the hematopoietic tissues. The results of this work were transmitted in Special Report No. 39, which is included herein. This report describes the tests made in detail. The conditions found in autopsies are recorded and discussed. Charts showing the various body conditions of the animals during the periods of observation, and photographs of internal sections taken in autopsies are also attached. The conclusions define the lethal dose of dichloroethylsulphide when injected intravenously, and the body conditions produced.


The much debated question of the effectiveness of Lewisite I as a war gas was studied from the medical standpoint. In general, tests were made in comparison with the effects obtained by a corresponding exposure to mustard gas.

An extensive study of this series is embodied in Special Report No. 38, which described a direct comparative examination of the lesions produced on horses' skin by parallel exposures to dichlorethylsulphide and Lewisite I. This report is included herein. It contains a full study of the histological changes in the lesions produced by dichlorethylsulphide through a period of 32 days from the initial exposure, being the time required for complete reparation of the injured tissues. The comparative study of Lewisite I lesions was not continued longer than the beginning of reparative changes, which was first observed at the end of 48 hours after initial exposure. In addition, the report carries with it photomicrographs of the lesions produced by each substance in various stages of development. The main conclusion from this work was, that necrosis is produced earlier by the Lewisite I, that the reaction is more intense and deeper, but, that reparative changes begin within 48 hours after exposure by Lewisite I. Reparative changes do not begin in the case of mustard gas until the end of the first week after an application of corresponding concentration.

Less extensive tests were made of the effects of Lewisite I on the skin of man. The result of these are noted in weekly reports.1 The results are not concordant with those obtained on horses' skin. The experiments were carried out with two types of application, namely, as alcoholic solution of equal concentrations, and as the saturated vapors of the two substances. In the first case, single drops of the alcoholic solution being applied to the skin, mustard gas in 0.2 percent concentration caused erythema but no blister, whereas 1 percent solution Lewisite I gave no reaction. Five percent solutions of Lewisite I caused erythema in one case and the formation of a small vesicle in another. However, when 10 percent solutions were used, the Lewisite I lesion was in the early stages decidedly more severe. The saturated vapors gave nearly equal effects in exposures of seven and one-half minutes. In the actual experiments four out of seven subjects showed the greater reaction to Lewisite I. It should be noted in this connection that the saturated vapor of Lewisite I at 20° contains slightly more than ten times the absolute concentration in milligrams per liter present in the saturated vapor of dichlorethylsulphide at the same temperatures.


The results obtained here on the action of Lewisite I on skin of man, especially those obtained with alcoholic solutions, accord, in general, with the results obtained by the British antigas department, and by the American University Experiment Station. The effects observed when summed up indicate that in minimal concentrations, mustard gas is much more effective in producing lesions. However, beginning with applications of the two substances in amounts per unit of skin area at which Lewisite I first shows effect on human skin, the reaction with Lewisite I increases very rapidly, and with high concentrations considerably exceeds that obtained with dichlorethylsulphide. The reason for this situation may lie in the fact pointed out by the Chemical Warfare Service Laboratory, that Lewisite I is very readily hydrolyzed as compared to dichlorethylsulphide. It may be supposed, accordingly, that small amounts of Lewisite I are hydrolyzed by moisture on the skin surface before penetration, but that Lewisite I, having once actually gained entrance through the skin, produces the more prompt and severe reaction.

A final study on Lewisite I was that of the effects of the intravenous injection into rabbits. Three of these animals were made the subjects of the experiment. One injected with 0.008 gram per kilogram body weight, died immediately. The second, injected with 0.005 gram per kilogram body weight, died three and one-half hours later with massive edema of the lungs. The third injected with 0.001 gram per kilogram showed loss of 100 grams in body weight, but appeared otherwise normal, and survived. This investigation was cut short by the closing of the field.


In order to be assured of safety in the use of this substance as a camouflage agent, it was necessary to determine the toxicity of the compound itself out of the presence of mustard gas. The proposed tactical method was described in the Summary Report of the Field Gas and Artillery Sections.2 The camouflage material would be employed on fronts where an attack was to be made within a short time. It was accordingly necessary that material to be used should be certainly nontoxic. The results of these tests were communicated in Special Report No. 57, which is included herein. Some reaction on the animals was observed. The tests, however, were severe. They were comprised of such orders of exposure as one-half to two hours in concentrations of 0.73 milligram per liter, and six hours in concentrations of 0.012 milligram, or 1: 500,000. If, however, it would be necessary to use only 10 percent of dimethyltrithiocarbonate in a shell filling to disguise the odor of mustard gas, as the tests described elsewhere have shown, it would not be desirable to use high concentrations in the "fake" mustard gas attacks.

The possible delayed effect of lung lesions followed by secondary infections from exposure to this gas was considered in the work described in Special Report No. 57. However, it appears that the animals upon which tests were made, were taken from a group of which a number were already affected with some form of bronchial trouble or pneumonia. Accordingly, the lesions in the bronchial tubes and lungs found in some of the animals autopsied could not be regarded as evidence of any value on this point. The closing of the field prevented determination of this point.



These autopsies were carried out by the Pathological Section for the greater part in connection with field experiments, yet in very considerable number for the chamber experiments on the Physiological Section. The total number of animals examined was roughly between 150 and 200 The examination and results obtained are recorded in every case with the report of the main investigation, of which each formed a part.




This work was commenced originally to supplement the American University experimentation by testing the possible usefulness of French species of snails for the detection of mustard gas. It was among the earlier studies by the medical sections of the field, and facilities were not available for full laboratory tests. However, the experiments and results recorded in Special Report No. 15 were obtained in response to a telegraphic request from the United States. This report is included herein. It was not found that these French species of snails or slugs could be used as field indicators for mustard gas.

Immediately following the work just considered, the possibility of the use of other species of low-form organisms was taken up. These considered were tadpoles, water insects (acilius sulcatus), and mosquito larva. These happen to be forms that were most easily obtained in the immediate vicinity. An experimental field chamber was constructed with capacity of 1 cubic meter. The animals to be tested were placed in a pail of water, and either subjected to a saturated atmosphere of the gas in the chamber, or to a direct addition of blue cross or yellow cross shell filling to the water.

The results of these tests were as follows; (a) Concentrations of 1 c.c. mustard gas to 10 liters of water caused the death of nearly all the forms tested in from 10 to 24 hours, but a pail kept in an atmosphere containing vapor did not affect the animals tested during a period of 24 hours. (b) Concentrations of 1 c.c. blue cross to 10 liters of water produced signs of distress in the animals tested in about half an hour's time, and all died within 12 to 24 hours. When 0.5 c.c. was used in 10 liters of water the same early signs of distress were noted, but there was no subsequent death in the 24-hour period of observation.

It was concluded that these animal forms do not offer sufficiently definite early symptoms in the presence of small yellow cross and blue cross concentrations of shell contents in water to be of much value as "indicators" in the field.

This same general problem was brought forward from time to time again, but no further investigations were made. The possibility of the use, however, of microorganisms as ameba or paramecium were considered. However, because of the difficulties of carrying on proper experimentation at the field, because no data indicated a probable successful outcome of such experiments, and finally, because it was not believed that any microscopic method would be acceptable to officers in the field, such investigations were not reopened.



In conjunction with the experiments of the chemistry section on the determination of concentrations in mustard gas and air mixtures, physiological tests were studied for the detection and even for the estimation of concentrations of such mixtures. The experiments on the use of sodium platinic iodide as an indicator paper and the animal tests carried out are mentioned in Summary Report of the Chemistry Section.3

A method of different sort was that proposed for the estimation of the composition of shell fillings, by the standardization of the reaction on horses' skin. Alcoholic solutions were utilized, made in dilution in known ratio from the original shell contents. The results obtained showed that the estimation might be made by this method to within 10 to 15 percent of the concentrations of mustard gas present in the alcoholic solution. These estimations were made on solutions unknown to the observer. However, rather grave discrepancies appear between the estimations of two different observers. This work was not made the subject of a special report, but was noted in a weekly report.4



A very thorough test was made of the special fabric suit proposed and constructed by the Research Division, Chemical Warfare Service. These tests are embodied in Special Report No. 44, which is included herein. This investigation comprises laboratory tests of the fabric as protection for horses' skin and human skin chamber tests against nominal concentrations and intensive field experiments. The results were in general very favorable to the suit. The protection against mustard-gas vapor, for which the suit was originally designed, was entirely adequate. Further tests with various severe exposures to liquid mustard gas demonstrate that it might not be depended upon too far for such protection. A notable feature of these experiments was the lengths to which the field experiments were carried, and resulted in rather serious effects with some of the subjects.


A few comparative tests were made on ointments for protection against mustard gas. These dealt for the most part with preparations already proposed. This work is embodied in Special Report No. 49, included herein. They are, in general, rather unfavorable as to the efficacy of such materials. Among the most important results noted are that these ointments tend rather to delay the development of the lesions than to moderate their ultimate severity, and that while some protection is afforded by sag paste, or hypochlorite ointment, it is rarely complete, and is entirely absent in a considerable portion of cases. It is also noted that horses' skin is not satisfactory for such tests, but that human skin must be used. Sag paste was considered preferable to a mixture of calcium hypochlorite with petrolatum, since it is nonirritating and probably more permanent than the latter.

An experiment was made at Hanlon Field on varying decompositions of protective ointments by the addition of a protein component. It had been reasoned that if dichlorethylsulphide entered into combination directly with


the tissue elements its vesicant powers might be counteracted in this manner. Protein and lipoid fractions prepared from animal tissues were used specifically.

A note on this work was transmitted,5 but no greater effectiveness of the ointments was indicated, and the work was not carried further.


The question of finding a material which would protect horses' feet against mustard gas was taken up at separate times in two brief studies. Neither succeeded in determining a satisfactory preparation.

In the first tests tar which had been boiled for three hours to drive off lower fraction was used. Experiments were made separately to test the protection of the skin of the forefoot and of the frog. The tests on the skin showed that, although immediately after cooling the tar had a smooth hard surface, within a half hour cracks appeared over the tar surface. These extended through the tar covering and exposed the skin underneath. It was seen that such fissures would form traps capable of retaining contaminated earth in contact with the skin. It was concluded that the preparation, on account of its brittleness would be worse than useless if applied for protecting the skin. On the other hand, tests on the hoof itself showed conclusively that no effect or result of lameness could be produced by the most severe exposures to mustard gas, regardless of protection. Accordingly, protection for the frog was unnecessary.

The second investigation was made on material specially prepared for the purpose at the Chemical Warfare Service laboratory. This new preparation was subject to essentially the same defect as the material tried first, though not quite to the same degree. When applied directly, fissures developed some hours later. It was then found that the cracking could be avoided by means of clipping the hair before applying. Yet in face of this latter precaution, in the course of 24 hours the mixture became so softened that it flowed down over the hoof.

The work was carried out and reported shortly before the closing of Hanlon Field. 6 It is not likely that work would have been taken up again since it had been found, on inquiry of the British, the French, and also the chief veterinarian, A. E. F., that no real need existed for protection of horses' feet against mustard gas.


Study of this question was begun very late and actually limited to a test on the effectiveness of potassium permanganate solution in the treatment of mustard gas burns. The results were conclusively unfavorable. The lesions were obtained by the exposure to the vapor of the purified distillate from German yellow cross shell contained in 4-mm. glass tubes at room temperatures. The tubes contained, at the end opposite from that applied to the skin, cotton moistened with dichlorethylsulphide. The treatment consisted in swabbing with 1 percent solution of potassium permanganate. In each case four exposures were made, two of which were treated with permanganate solution and two left as controls. In the series of six experiments the treatment was commenced at various periods after exposure and repeated at recorded intervals. It was found that in no case did the treatment alter the course of the mild lesions caused by controlled exposure to mustard-gas vapor.7



A very considerable number of human cases of gas poisoning were studied at certain of the Army hospitals, specifically Evacuation Hospitals No. 1 and No. 2, Base Hospitals No. 15, Nos. 45, and No. 58, and the Justice group of hospitals. Also pathological specimens were received in cases where death occurred. This represented largely deaths from mustard gas and from second- ary infection of the respiratory tract. However, except in few cases in the weekly reports, this work was not reported through Hanlon Field, and consequently can not be included here.



1. There exists evidence pointing toward the general toxicity of dichlorethylsulphide, both when administered by inhalation and when injected subcutaneously and intravenously.
(a) Lynch 8 found that dogs gassed with high concentrations of dichlorethylsulphide--0.3 mgm. per L. for one hour--exhibited symptoms similar to those produced by injection, and not referable to the primary irritation of the respiratory tract. These symptoms were salivation, vomiting, bloody diarrhea, hyperexcitability and convulsions, slow, irregular pulse, becoming rapid before death, and attributed to vagal paralysis. Furthermore, the absorption of dichlorethylsulphide during inhalation was shown by the appearance of the hydrolysis product, dihydroxyethylsulphide, in the urine.
(b) Severe symptoms, ending usually in death within 24 hours, followed the intravenous injection of approximately 0.01 gm. per kilo in dogs. At autopsy, intense congestion, and often extensive hemorrhage into the intestinal mucosa, was the only lesion noted.9
(c) Changes in the formed elements of the blood after injection of Yperite have been recorded by Muratet and Fauré-Fremiet,10 and by Jolly,11 and may be taken as further evidence of the systemic toxicity of this substance.

2. The following report deals with the effects of the intravenous injection of dichlorethylsulphide in rabbits. Special attention was given to the alterations in the blood picture and in the blood-forming organs, since these were the most striking of the results observed to follow the administration of this toxic substance when introduced directly into the blood-stream.



(a) Difficulty was at first experienced in obtaining suitable emulsions for injection. In the earliest experiments the given amount of a 10 percent alcoholic solution by weight of dichlorethylsulphide was suspended in 0.85 percent salt solution, shaken vigorously and immediately injected, after allowing the larger globules to settle out. The dose administered was therefore con- siderably less than the total amount taken. Later it was found that a satis- factory suspension could be made by emulsifying in 30 percent alcohol in distilled water. A slightly milky emulsion was obtained which, after shaking, did not separate out in the time necessary for injection. The suspension was


prepared from a recently made and accurately weighed 10 percent solution in absolute alcohol and immediately injected to avoid hydrolysis. The dichlorethylsulphide used was a distillate from a German yellow cross shell content and was actively vesicant.


No attempt was made to determine accurately the minimal lethal dose. It was found in the earlier experiments that the limit of tolerance was in the neighborhood of 0.01 gm. per kilo body weight, and in the last six rabbits used a uniform dose of 0.005 gm. per kilo was chosen.


1. Emaciation and loss of weight was noted in all the rabbits which survived two days or more after the injection. In all but one animal (rabbit No. 12, loss of weight only 100 gms.) this was associated with diarrhea, and perhaps was due to it.
2. Nervous symptoms.- These were observed only in rabbits No. 4, No. 9, and No. 13, dying one hour, one and one-half hours, and during the night following the injection. The animals showed extreme restlessness, incoordinate movements, retraction of the head, transient spasticity, but no definite paralyses or convulsions.
3. Respiratory symptoms.- No definite or characteristic respiratory symptoms were observed, even in animals dying a few hours after the injection, in which the occurrence of pulmonary embolism might have been suspected.
4. Diarrhea occurred in six animals, in all but one associated with the finding of gross lesions of the intestinal tract at autopsy. The feces were copious, fluid dark brown, not grossly admixed with blood or mucus.
5. Edema of the ears of wide extent invariably followed injection, even when great care was taken to avoid introducing the material outside the vein. It would appear that the dichlorethylsulphide diffuses readily through the vessel wall. It was also observed that after the injection was begun an area of blanching involving the neighboring skin over a width of several centimeters at once appeared. This persisted for a few minutes after the injection, after which the normal circulation was restored.



(a) Of 12 rabbits injected, 4 showed definite pulmonary lesions (Nos. 1, 4, 9, 13). These animals all died or were killed within a period of from 1 hour to 22 hours following the injection, and none of the 8 animals surviving over 24 hours has shown significant gross or microscopic lesions.
(b) The changes observed were: (1) Irregular area of edema, in part fibrinous; (2) areas of atelectasis and emphysema; (3) accumulations of leucocytes in the capillaries, often showing caryorrhexis and fragmentation; slight emigration into the alveoli. The trachea and bronchi were normal, except that they contained a homogeneous coagulum. No thrombi were found in the capillaries or larger vessels. The small pulmonary arteries were thick walled and appeared contracted; clear vacuoles were seen beneath the endothelium. The significance of this finding is not clear, as somewhat similar pictures may be seen in normal animals.


(c) It was of course not possible to conclude from the histological findings that dichlorethylsulphide was eliminated by the pulmonary epithelium; on the other hand, no support was found for the view that the lesions were the result of capillary embolism due to impaction of dichlorethylsulphide globules. The fact that animals which survived for a longer period showed no pulmonary lesions might suggest that the edema was the result of a direct and immediate action of the dichlorethylsulphide upon the pulmonary capillaries.


(a) No lesions were found in the esophagus. Rabbit No. 1 showed hemorrhages into the pyloric portion of the stomach and in the duodenum. Of the remaining rabbits, three (Nos. 20, 24, and 26) had a severe diphtheritic enteritis affecting the middle or lower portion of the small intestine. Rabbit No. 20 showed also patches of membranous inflammation in the large intestine. Rabbit No. 2 also had a diphtheritic colitis, but this proved to be coccidial in origin. Seven rabbits were free from lesions of the intestinal tract.
(b) The liver showed no significant changes.


(a) No detailed study was made. Capillary thrombosis or hemorrhage were not found.


(a) Definite changes were present in the kidneys of only rabbits No. 1 and No. 4. The capsular spaces contained hyaline globules and occasionally red blood cells; hyaline material was also present about the blood vessels in the intermediate zone between cortex and pyramids. Blood cells and hemoglobin (?) casts were found in the collecting tubules. The urine was not examined.



The observations of previous workers upon the blood changes following the administration of dichlorethylsulphide having yielded somewhat conflicting results, it was decided to undertake a detailed study of a small series of rabbits, eliminating as far as possible incidental factors which might influence the blood picture. Of these incidental factors, variations in the surrounding temperature were found to be the most disturbing. When rabbits were exposed to a temperature of from 40° to 50° C. for one-half hour, the total leucocyte count showed a tendency to fall, although there were individual exceptions. Counts made shortly after removing the animals from the warm chamber to room temperature (15° to 20°) invariably showed an abrupt and striking rise, as shown in the table and charts (infra). After this point had been established, the rabbits during the observational period and following the injection were kept in the laboratory to avoid sudden chilling or abrupt temperature change. Daily counts were made at approximately the same time (9 to 11 a. m.), before food was given. The differential counts are based upon an enumeration of 500 cells, except where the extreme leucopenia made this impracticable. In spite of the usual precautions, unexplained variations occurred which made it


difficult to draw conclusions as to the percental fluctuation. In making the counts, blood was always taken from the uninjected ear. Where both ears had been used, a small cut was made in the skin of the abdomen, and blood taken from one of the superficial abdominal veins. At least two daily counts were made on each rabbit before injection.


Stress of other work prevented a detailed study of the numerical variations in the erythrocytes. Preliminary observation (rabbit No. 2) showed no significant change. There were no striking morphological changes pointing to a marked anemia, except perhaps the occurrence of a moderate anisocytosis in the terminal stages. Polychromatophilia was not infrequently seen in smears from normal rabbits. Rabbit No. 27, during the period in which recovery from the effects of an injection was taking place, showed numerous normoblasts--an indication that the erythroblastic tissue had suffered injury, as well as the leucoblastic. Histological study of the bone marrow and spleen afforded further evidence of the toxic action of dichlorethylsulphide upon the formation of red blood cells, which will be described.


In all but 2 rabbits (Nos. 25, 26) of those surviving more than 24 hours, there occurred after a single injection of 0.005 to 0.01 gm. of dichlorethylsulphide per kilo, a pronounced fall in the number of circulating leucocytes. In these 2 rabbits, a second injection of the same dose after seven days, was followed by the typical reaction. The leucopenia was preceded by a transient rise in only 1 rabbit (No. 24), but it is only fair to state that no counts were made at short periods following the injection. Although degenerated leucocytes with poorly staining and fragmented nuclei and vaculated cytoplasm were occasionally found in smears, they were infrequent, and even in the presence of an extreme leucopenia the rare leucocytes present in the films were usually normal morphologically. (The data showing the alterations in the blood count are presented in Table 95, and in Charts XXXII to XXXVII.) As regards the behavior of the different types of leucocytes, a study of the data shows the following:

1. In some cases, the injection was followed by an absolute and percental increase in the polymorphonuclears, which fell rapidly with the onset of the leucopenia. In the terminal stages the polymorphonuclears practically disappeared from the peripheral blood.
2. In other cases, an initial fall in the number and percentage of polymorphonuclears was followed by an absolute and relative increase. This secondary rise was associated with the appearance of unripe forms in considerable numbers (rabbits Nos. 25, 27), and coincides with regenerative activity of the bone marrow, as shown by a study of sections from rabbits killed at this stage.
3. The leucopenia was accompanied by a relative lymphocytosis. The absolute number of lymphocytes was diminished in the later stages, and lagged behind the granular cells in cases in which regeneration was occurring.
4. There was a percental increase in the large mononuclear cells, but their absolute number was unchanged or diminished.



Blood platelets were found at all stages, and showed no alterations.


1. Although the appearances of the marrow varied in different animals, there was very clear evidence of the destructive effect of the dichlorethylsulphide upon the blood-forming elements. The variations observed could be correlated with varying stages of injury and repair, and these again were reflected more or less closely in the blood picture during life. The following brief descriptions will illustrate the different phases observed.
2. Rabbit 12: Killed four days after injection. At time of death, the leucocyte count had fallen to 400, of which 46 percent were polymorphonuclears. The marrow of the femur contained a large amount of adipose tissue, the fat cells being separated by a loose edematous tissue containing less than the normal number of cells. The leucocytes of the granulocyte series were almost without exception degenerated. The cytoplasm in sections stained with Wright's stain (normal control) showed no granules. The nuclei stained diffusely and were frequently fragmented. The megacaryocytes also showed degenerative changes. Some of them contained clumps of pink staining, hyaline material in their cytoplasm. There were islands of apparently normal erythroblasts. The blood sinuses were wide and intensely congested with normal-appearing red blood cells; they contained practically no nucleated elements. The appearances were interpreted as indicating an active injury to the bone marrow. The toxic action seems to have affected especially the granular cells.
3. Rabbit 26: Died three days after a second injection of dichlorethylsulphide. Leucocytes on last two days had fallen to 900 and 800, respectively, a smear of the peripheral blood showing very few nucleated cells, and these almost exclusively large and small mononuclears. Sections of the marrow show an extreme aplasia, comparable with that seen in experimental benzol poisoning. Myelocytes, polymorphonuclears, and megacaryocytes have practically disappeared. There are loose collections of normoblasts scattered through the edematous tissue. Occasional globular fragments of chromatin, often inclosed in phagocytes, represent the remains of the destroyed cells. The sinuses are congested and contain no leucocytes. (See fig. 233.)
4. Rabbit 25: Killed on the seventh day following a second injection. The leucocytes, which had fallen to 1,000 on the third day after injection, then rose to 1,400 and 2,100, and at the same time large numbers of myelocytes appeared in the peripheral blood. The histological picture shows the effect of a previous injury and at the same time an active regeneration. The predominant cell type is the myelocyte, the granules of which are definite and well stained in Wright preparations. There is also a fair number of adult polymorphonuclears. The myelocytes are congested in islands, as are also the erythroblasts. Megacaryocytes are numerous and are not degenerated. There are many mitoses. The marrow on the whole is less cellular than normal marrow, and new fat cells are in process of formation. There is much hemosiderin pigment, chiefly intracellular, which may be taken as evidence of previous blood destruction.


5. Rabbit 27: After the first injection, there ensued a fall in the leucocytes from 37,000 to 2,700 on the fourth day, after which there apparently occurred an active regeneration. The count rose again to 32,400, followed by a slight drop to 25,400. At this point a second injection was given, which again was followed by a marked leucopenia, the leucocytes falling to 300 per c. mm., at which point the rabbit was killed. The marrow histologically showed evidence of an initial injury, followed by repair and a fresh destruction of the regenerating cells. There were large areas of almost complete aplasia, similar to that described in rabbit No. 26. But there were also hyperplastic areas, composed of aggregates of large cells with poorly staining nuclei, which under the high power were found to be distorted and obviously degenerating. It would seem that these hyperplastic foci have been again injured by the second injection of the toxic substance.


1. Changes in the follicles--fragmentation of lymphoid cells with phagocytosis of chromatin particles--were seen in only two of the rabbits (Nos. 1 and 13). (See fig. 234.) Both of these animals died within less than 24 hours after injection. In the remaining rabbits the follicles were normal, or at least showed no signs of either active destruction or excessive proliferation.
2. The sinuses in the majority of the rabbits contained large mononuclear cells laden with blood pigment. The most striking change, however, was the paucity of free cells in the meshes of the reticulum. The sinuses were separated by strands of cells with pale oval nuclei, evidently belonging to the reticular elements.


1. The two rabbits (Nos. 1 and 13) which showed acute destruction of the lymphoid cells of the splenic follicles also showed cytolysis of the lymphocytes of the thymus and of the intestinal lymphoid tissue.
2. Because of the great susceptibility of the small thymus cells and of the tissue lymphocytes generally to various injurious agents, this was not regarded as a specific effect of the dichlorethylsulphide, particularly as it was not present in those cases in which the destruction of the bone-marrow elements was extreme.
3. In experiments with suspensions of thymus cells in vitro, it was found that the permeability of the cells to trypan blue, which may be taken as a sensitive index of injury, was not materially increased by the addition of moderate concentrations of dichlorethylsulphide (less than 1 percent).


1. Of the effects of dichlorethysluphide when introduced intravenously in rabbits, the most interesting is that upon the blood-forming organs, and upon the leucocyte content of the peripheral venous blood. Within three or four days there may be produced by a single small dose a practically complete exhaustion of the marrow, of granular cells of all types, and a corresponding disappearance of these elements from the blood stream. If the animal survives the initial injection, regeneration occurs and the normal-blood picture is restored. A second injection will cause a renewed injury to the bone marrow.


2. This striking reaction has attracted the attention of other observers. Muratet and Faurè-Fremiet 10 reported upon blood examinations in a series of 6 rabbits, 4 of which were poisoned by inhalation and 2 by subcutaneous injection. In all the animals the following changes were noted: The red blood cells were augmented in number soon after exposure (increase to 6,000,000 to 10,000,000), and then gradually diminished. Nucleated forms were sometimes seen. There was no polychromatophilia or other degenerative change. The leucocytes showed a rise following the intoxicating dose, but later diminished progressively. There was a relative lymphocytosis. Morphologically, degenerative changes were found in the leucocytes. The polymorphonuclears showed abnormal lobulation of the nucleus, dissolution of the granules, and finally a breaking up of the chromatin into spherical masses. The lymphocytes also showed degenerative changes, and many abnormal cells which could not be identified were present in the films.
3. Jolly 11 repeated the experiments of Muratet and Faurè-Fremiet 10, using both dogs and rabbits, and administering the dichlorethylsulphide by both inhalation and subcutaneous injection. His results were less consistent than those of the previous workers, but a study of the figures obtained seems to show a rather constant diminution of leucocytes after injection; with inhalation. the results were variable.
4. Zunz 12 observed, in severely gassed human cases, a leucopenia with relative lymphocytosis developing after several days.
5. The similarity of these effects to those produced by benzol bring up the question as to whether it was the dichlorethylsulphide itself, or the chlorbenzene solvent passing over in small amounts into the distillate, or possibly one of the impurities contained in the dichlorethylsulphide, which was responsible for the injury to the hematopoietic tissues. It was hoped to carry out further experiments to determine these points; but it does not seem probable that the very small amount of chlorbenzene which would be contained in the dose of dichlorethylsulphide administered in these experiments would suffice to cause so severe an injury.c
6. Another interesting feature of the pathology of systemic dichlorethylsulphide poisoning was the frequent occurrence of severe lesions of the intestine. These suggested an elimination of dichlorethylsulphide or its products through the intestinal mucosa, since a direct action after intravenous injection was highly improbable.
7. The significance of the lung lesions has already been discussed. They resembled those caused by gases of the "suffocative" type rather than those produced by inhalation of dichlorethylia. Whether an elimination takes place through the lung was not determined.


1. The lethal dose of dichlorethylsulphide (distillate from German yellow cross shell) when injected intravenously into rabbits, is from 0.005 to 0.01 gm. per kilo body weight.
2. Rabbits dying within 24 hours showed extensive hemorrhages and edema of the lungs.
3. Severe lesions of the intestinal tract were present in about one-third of the rabbits.

c Subsequent experiments carried out at the Brady Laboratory, Yale University Medical School, by one of the writers of the report showed that the leucotoxic effect was not due to admixture of chlorbenzene in the shell fillings. (JournalExperimental Medicine, 1920, xxxi, No. 1, 71.)


4. Dichlorethylsulphide, injected intravenously, is specifically poisonous for the hematopoietic tissues. Severe lesions were caused in the bone marrow and the number of circulating leucocytes was markedly diminished. In animals surviving the injection, regeneration occurred. The granular cells of the bone marrow seem to be more sensitive than the lymphoid cells and the erythrocytes.


RABBIT 1 (serial No. 6).-Healthy animal; weight, 2,030 grams before exposure. August 30, 1918, at 12.30 p.m. received intravenous injection of dichlorethylsulphide in marginal vein of ear.
Dose.-One-hundredth cubic centimeter of pure dichlorethylsulphide, or 0.005 c.c. per kilo of body weight, in the form of a 10 percent alcoholic solution freshly prepared, 1 c.c. of which was added to 4 c.c. of normal saline.
August 20, 4.45 p. m., seemed very weak and apathetic; 7 p. m., still very weak; ears and body very cold; developed a profuse watery diarrhea. August 21, 10.20 a. m., still very weak but seems much better. Killed by a blow on the back of the neck. Autopsy at 11 a. m.
Gross findings.-Respiratory tract: Injection of submucous venules of trachea. Lungs show some interstitial and irregular dark red areas scattered through all lobes. Gastrointestinal tract: Fresh submucous hemorrhages in the pyloric end of the stomach. For a distance of 5 to 6 cm. below the pyloric end, the duodenal wall shows numerous fresh hemorrhages. The rest of the intestinal tract is normal. Other organs normal.
Microscopic findings.-Trachea is normal. Lungs: Section shows irregular congestion of alveolar wall with an increased number of polymorphonuclear leucocytes in capillaries and alveolar lumen. In the congested areas the alveoli are partially collapsed. The nuclei of the leucocytes show marked pycnosis and fragmentation. There are patchy areas of edema, sometimes with a little fibrin. In other places there are hemorrhages into the alveolar lumen. In one slide, there is a distinct area of bronchopneumonia; a vessel at the apex of this area contains a fresh thrombus. The bronchi are generally normal except that they often contain red blood cells and coagulated edema fluid. Stomach: There are several small areas of focal hemorrhagic necrosis with beginning leucocytic reaction. Duodenum: The glandular tubules are separated in many places by fresh hemorrhages. The leucocytes in the extravasated blood are pyenotic and fragmented. The surface epithelium in places is necrotic. Spleen: The follicles are largely replaced by large reticular cells with pale nuclei and indefinite cell outlines. The small lymphocytes at the margin of the follicles are pyenotic and fragmented. Large phagocytes filled with chromatin fragments are found in pulp and lying in sinuses. Thymus: Cytolysis of the small thymus cells; nuclear fragments inclosed in phagocytes. Epithelial components of gland unchanged. Kidney: Globular coagulum in the capsular spaces. Changes not very marked.

RABBIT 2 (serial No. 15).-Young healthy animal. Weight before exposure, 2,125 grams. August 20, 1918, at 12 m., received intravenous injection of dichlorethylsulphide in marginal vein of ear.
Dose.-Five one-hundreths cubic centimeter, or 0.025 c.c. per kilo of body weight, in the form of a freshly prepared 10 percent alcoholic solution diluted with normal saline.
Injection produced no immediate effect. Rabbit showed progressive emaciation. Died August 25, at 11 a. m. Autopsy 15 minutes after death. Body weight, 1,760 grams.
Gross findings.-Emaciation. Moderate pulmonary emphysema. Gastrointestinal tract: Diphtheritic colitis of large intestine. Spleen: Flat and soft.
Microscopic findings.-Lungs: Emphysema present, with some fresh alveolar hemorrhage. Spleen: Follicles normal. The red blood cells are practically in pulp reticulum; none in sinuses. Sinuses contain large multinuclear cells with abundant cytoplasm, which contain much brown pigment.

RABBIT 4 (serial No. 10).-Healthy animal. Weight before exposure, 2,025 grams. August 22, 1918, at 3.20 p. m., received intravenous injection of dichlorethylsulphide in marginal vein of ear.
d The charts include only those animals in which blood counts were made during a preliminary observational period.


Dose.- Five one-thousandths cubic centimeter per kilo of body weight, given in a 10 percent alcoholic solution diluted with normal salt solution.
At 4.30 p. m., rabbit was found in convulsions, showing opisthotonos, spastic paralysis of hind legs, with occasional clonic spasms, rapid dyspnea, etc.; 5.20 p. m., died.
Gross findings.- Organs appear normal aside from hemorrhagic areas in the lungs.
Microscopic findings.- Lungs: Patchy areas of edema and irregular congestion. There is no true bronchopneumonic process present. Spleen: Normal follicles; no cytolysis of lymphoid cells. Sinuses not congested, nor do they contain abnormal elements. Other organs are not markedly abnormal .

RABBIT 9 (serial No. 41).- Healthy female. Weight before exposure, 2,120 grams. August 30, 1918, received intravenous injection of dichlorethylsulphide in marginal vein of ear.
Dose.-Five one-thousandths cubic centimeter per kilo of body weight, given in a freshly prepared 10 percent alcoholic solution diluted with saline.
No symptoms followed this injection. September 2, 1918, 10 a. m., injection repeated; 11.30 a. m., animal began to show convulsive movements, with periods of rest during which she lays on her side; diarrhea; 2 p. m., killed.
Gross findings.-Frothy discharge from nose and mouth; frothy fluid in esophagus and trachea. Lungs show areas which are sunken and dark red in color. Other organs normal.
Microscopic findings.- Lungs: Patchy areas of edema, some part of coagulum being definitely fibrous. The capillaries are congested and crowded with leucocytes. The nuclei of these are very often pycnotic and caryorrhectic. Bronchial epithelium is normal; submucous edema and leucocytic infiltration is present in some of the bronchi. (Fig. 229.)

FIG. 229.- Rabbit 9, serial No. 41. Death 2 ½ hours after second injection of dichlorethylsulphide. Lung: Edema (oe) in part fibrinous. Stasis of leucocytes (Lc) in capillaries. Coagulum in bronchus (B) and distension of peribronchial lymphatic vessel (L)

RABBIT 12 (serial No. 97).-Healthy half-grown female. Weight before exposure 1,620 grams. September 16, 1918, at 2.30 p. m., received intravenous injection of dichlorethylsulphide in marginal vein of ear.
Dose.-One-hundredth cubic centimeter of dichlorethylsulphide per kilo of body weight in the form of a 1 percent solution in 30 percent alcohol and distilled water freshly prepared. Progressive emaciation and weakness until death. September 20, 1918, at 3 p. m., killed by blow on back of the neck. Autopsy September 20, at 3.15 p. m. Weight at autopsy 1,520 grams.
Gross findings.-Normal. Intestinal tract normal. Rest of the organs are practically normal.
Microscopic findings.-Lungs and gastrointestinal tract, normal. Spleen: Follicles normal. Atrophy of the pulp tissue. Large spindle-shaped cells occur in the walls of the sinuses. Bone marrow: Fat cells are present in some numbers, and the fatty tissue shows edema. The marrow is much depleted of its cells. There are no normal polynuclears and no granulated myelocytes. The cytoplasm of the megacarvocytes shows some degenerative changes. Mitoses are rare. Fragmentation of cells is fairly common. There are numerous nucleated red blood cells. The vessels are congested and contain many nucleated cells.

RABBIT 13 (serial No. 55).-Healthy female; half-grown; weight before exposure, 1,660 grams. September 16, 1918, at 2.30 p. m., received 0.016 c.c. of dichlorethylsulphide, or 0.01 c.c. per kilo of body weight. This was given as a 1 percent solution in 30 percent alcohol and distilled water freshly prepared. Rabbit was accidentally choked while being


injected, but she speedily recovered. Later, convulsive movements developed, which were incoordinate rather than paralytic. Died, September 17, early in morning. Autopsy, September 17, at 8.15 a. m. Weight after death, 1,550 grams.
Gross findings.-Lungs appear to be edematous. Other organs are normal.
Microscopic findings.-Lungs: Patchy edema; irregular areas of congestion where the alveoli are partially collapsed. Many polynuclear leucocytes in capillaries or even in the alveolar lumen. The bronchi contain a small amount of edematous fluid but are otherwise normal. Spleen: In pulp, scattered cells show fragmentation, caryorrhexis and some pigment containing cells. Follicles show extreme destruction of lymphoid cells. Active phagocytosis in center of follicle of the fragmented cells by large mononuclear cells with abundant cytoplasm. Thymus: Destruction of lymphoid cells in the center of the lobule and phagocytosis by large mononuclear cells. (Fig. 230.)

RABBIT 20 (serial No. 96).-Healthy female rabbit. Weight before exposure, 1,850 grams. September 16, 1918, at 2.30 p. m., received intravenous injection of dichlorethylsulphide in marginal vein of ear.

FIG. 230.-Rabbit 13, serial No. 55. Died in less than 18 hours after intravenous injection of dichlorethylsulphide, 0.01 g. per kilo. Spleen showing caryorrhexis of cells of follicles with phagocytosis of chromatin fragments

FIG. 231.-Rabbit 20, serial No. 96. Died 4 days after injection of 0.1 gm. per kilo intravenously. Small intestine: Complete hemorrhagic necrosis of mucosa (M); extreme fibrinous edema of submucosa (Oe)

Dose.-One-hundredth cubic centimeter of dichlorethylsulphide per kilo of body weight in the form of a 1 percent solution in 30 percent alcohol and distilled water freshly prepared. Lost weight steadily from date of injection. Weight September 19, 1,550 grams Diarrhea, very profuse, noticed on September 19. This persisted until death. Died, September 20, at 10 a. m. Autopsy.-Weight before autopsy, 1,485 grams.
Gross findings.-Lungs normal. Gastrointestinal tract normal. The lower 30 cm. of the small intestine and patchy areas in the cecum are the seat of a necrotic enteritis. Other organs are normal.
Microscopic findings.-Lungs are normal. Small intestine: Partial to complete necrosis of intestinal epithelium. Numerous thromboses in capillaries of submucous layer. Intense submucous edema. Spleen: Follicles are composed of normal cells showing no fragmentation. Pulp contains very few cells and shows swollen reticula. Sinuses contain many large mononuclear cells with brown pigment in cytoplasm. Bone marrow: Shows many adult fat cells between which are numerous islands of myelocytes, polynuclear leucocytes, and nucleated red blood cells. The blood sinuses are congested, but contain very few nucleated cells. In the marrow, the number of megacaryocytes is normal. A few mitotic figures can be seen; some of the cells are fragmented, but there is no phagocytosis or pigment formation. The marrow presents evidence that it has suffered recent injury with early regeneration. (Fig.231.)


CHART XXXII- Total leucocyte counts, showing fluctuations due to experimental changes of outside temperature. Counts plotted by hour of day, space after each date representing 24 hours


RABBIT 23 (serial No. 141).- Healthy adult female. Weight before exposure, 2,000 grams. September 23, 1918, at 10 a. m., received intravenous injection of dichlorethylsulphide in marginal vein of ear.
Dose.-Five one-thousandths cubic centimeter per kilo of body weight, given in a 1 percent solution in distilled water and 30 percent alcohol freshly prepared. Progressive weakness and emaciation from day of injection. September 26, weight 1,900 grams. Diarrhea on September 27. Died, September 28 at 6 a. m. Autopsy. September 28, at 8.15 a. m. Body weight at death, 1,813 grams.


Gross findings.- Lungs: Normal. Intestinal tract presents no lesion, but the small intestine is full of an abundant mucoid secretion, while the large intestine is full of mushy unformed fecal material of a brownish color.
Microscopic findings.-Spleen: There are no changes in the follicles. The sinuses are distended and there are many large mononuclear cells containing pigment. Bone marrow: There are not many fat cells present. Polynuclear leucocytes, myelocytes, and megacaryocytes, however, are very abundant. (Fig. 232.) There are not many nucleated red blood cells. There are a few mitoses; some fragmentation can be seen.

RABBIT 24 (serial No. 104).-Healthy rabbit. Weight before exposure, 1,770 grams. September 23, 1918, at 10 a. m., received 0.005 c.c. of dichlorethylsulphide per kilo of body weight as an intravenous injection in the marginal vein of the ear. It was given in a 1 per cent soluition in 30 per cent alcohol and distilled water freshly prepared. Progressive weakness and moderate emaciation up to day of death. September 26, weight, 1,715 grams. Diarrhea just before death. September 27, at 7.30 a. m., died. Autopsy at 8.15 a. m., September 27.


Gross findings.-Lungs: Normal. Gastrointestinal tract: Middle two-fourths of small intestine seat of a necrotic enteritis. Contents are a thin and watery mucus. The wall of the large intestine is normal, but the contents consist of much abundant mushy fecal material. No formed feces present.
Microscopic findings.-Intestine: A necrotic membrane covers a submucous layer from which the epithelium has been desquamated. The underlying tissues are congested and infiltrated with various inflammatory cells. The epithelium which remains in the comparatively healthy areas is normal. Clumps of bacteria are present on the mucous surface. The centers of the Peyer's patches show a few areas of cellular destruction and caryorrhexis. Spleen: There is no destruction of cells in the follicles. Some of the follicles show areas which are entirely composed of epithelioid cells. The sinuses are distended with red cells, among which are large phagocytes filled with pigment. Between the sinuses are only cells of the reticular type. Bone marrow: The marrow is composed of many fat cells, numerous polynuclear leucocytes, granular myelocytes and islands of nucleated red cells. The cytoplasm of the megacaryocytes

FIG. 232.-Normal rabbit. Bone marrow of femur. Megaearyocyte (M)

shows some degenerative changes. There is, however, no fragmentation, phagocytosis, edema or pigment formation. (Fig. 233.)

RABBIT 25 (serial No. 168).-Healthy rabbit. Weight before exposure, 1,800 grams. September 23, 1918, at 11 a. m., received intravenous injection of dichlorethylsulphide in marginal vein of ear.
Dose.-Five one-thousandths cubic centimeter per kilo of body weight in a 1 percent solution in distilled water and 30 percent alcohol, freshly prepared. No symptoms followed this injeetion. Weight, September 26, 1,800 grams. October 1, injection repeated at 11 a. m. Progressive loss of weight followed. October 4, weight 1,580 grams. October 7, at 4 p. m. killed by a blow on back of head. Autopsy at time of death. Weight at death, 1,450 grams

FIG. 233.- Rabbit 24, serial No. 104. Died 3 days after second injection of dichlorethylsulphide, 0.005 gm. per kilo intravenously. Bone marrow of femur, showing extreme aplasia. The sinuses (B) are wide and congested. In the edematous pulp (oe) are many fat cells (F). Myelocytes. polymorphonuclears, and megacaryocytes are absent. There are few islands of normoblasts
The pale nuclei (R) are those of the reticular cells. (Somewhat lower magnification than Figure 232.

Gross findings.-Organs are apparently normal.
Microscopic findings .- Lungs: Slight emphysema and areas of patchy congestion. Spleen: Follicles are normal. Many large pigment cells in the sinuses. Very few wandering cells in sinuses. Bone marrow: Numerous embryonic fat cells, numerous islands of myelocytes, nucleated red cells and megacaryocytes, and numerous mitoses, present. There are not many polynuclear leucocytes. Pigment cells, however, are numerous. The marrow is actively regenerating.


RABBIT 26 (serial No. 174).-Healthy adult male. Weight before exposure, 2,045 grams. October 1, 1918, at 11 a. m., received intravenous injection of dichlorethylsulphide in the marginal vein of the ear.
Dose.-Five one-thousandths cubic centimeter per kilo of body weight given in 1 percent solution in distilled water and 30 percent alcohol freshly prepared. No symptoms followed first injection. October 7, at 11 a. m., injection repeated. Body weight, 2,050 grams. Progressive emaciation followed, with weakness. Profuse diarrhea occurred on October 10. Died, October 11, at approximately 6 a. m. Autopsy, October 11, at 8.30 a. m. Body weight at death, 1,860 grams.
Gross findings.-Gastrointestinal tract: Stomach normal. Small intestine: Lower 25 cm. of ileum is the seat of a necrotic enteritis. Large intestine is filled with fluid fecal material. The wall is normal. Other organs are normal in the gross.


Microscopic findings.-Small intestine: Necrotic pseudomembrane, polynuclear infiltration of submucous layer, and congestion of blood vessels. Desquamation of intestinal cells occur in some areas. Spleen: The pulp is depleted of cells, the follicles are atrophic. There are numerous large pigment cells in the sinuses. Bone marrow: Embryonic fat cells are present through the entire marrow. The marrow is entirely depleted of cells, except for a few islands of normoblasts. A few cells are fragmented, and occasionally some phagocytosis is seen. Myelocytes, polynuclear leucocytes, and megacaryocytes are generally absent. The marrow has been subjected to a marked destructive change.

RABBIT 27 (serial No. 177).-Healthy adult female. Body weight before exposure, 1,680 grams. October 1, 1918, at 11 a. m., received intravenous injection of dichlorethylsulphide in the marginal vein of the ear.
Dose.-Five one-thousandths cubic centimeter per kilo of body weight given in 1 percent solution in distilled water and alcohol freshly prepared.
After this injection the animal became somewhat weaker. Body weight October 7, 1,628 grams. Gradual recovery. October 7, at 11 a. m., the injection was repeated. Animal again became much weaker and somewhat emaciated. October 10, weight 1,605 grams. October 11, at 11 a. m., killed by blow on head. Autopsy at time of death.
Gross findings.-Bone marrow rather grayish in color. Other organs normal.





Microscopic findings.-Spleen: Follicles are normal. Pulp contains very few wandering cells. The sinuses are crowded with huge mononuclear cells with abundant cytoplasm which contain loads of brown pigment. Bone marrow: Part of the bone marrow is depleted of all marrow cells and the fatty tissue is edematous. In another area which is hyperplastic there


are quite a few cells present, including many abnormally lobulated leucocytes, and a few islands of nucleated red blood cells. Scarcely any of the normal bone marrow cells are present. There are a few megacaryocytes and some pigment-containing cells. The peculiar picture is probably due to the toxic action of the second injection on a marrow which is actively regenerating.


TABLE 95.- Alterations in blood count following intravenous injections of dichlorethylsulphide


TABLE 95.- Alterations in blood count following intravenous injections of dichlorethylsulphide - continued.


TABLE 95.- Alterations in blood count following intravenous injections of dichlorethylsulphide - continued


TABLE 95.- Alterations in blood count following intravenous injections of dichlorethylsulphide - continued





(1) Recently prepared 10 percent alcoholic solutions of the distilled contents of a Yellow Cross German shell, were applied to the shaved skin of horses over areas of 7 mm. in diameter, 0.1 c.c. being introduced into a glass tube held against the skin and allowed to evaporate. A few lesions resulting from other concentrations and methods of application were studied, but, in general, this procedure was followed in order to obtain comparable lesions at different intervals. When cocaine was used for excision the skin was infiltrated in a circle at a distance from the affected areas. The pieces taken included a seg- ment of adjacent normal skin.

(2) The tissue was fixed in Zenker's fluid, embedded in paraffin, and stained with hematoxylon-eosin, Van Gieson, Gram-Weigert, and by Weigert's method for elastic fibers. Considerable difficulty was experienced in obtaining satisfactory sections of the late stages, the dried necrotic tissue being extremely hard to impregnate and section.


(1) A few hours after exposure there develops a soft edema slightly more extensive than the area of application. The extent of the edema varies with the site of application, being greater in the loose tissue over the shoulders than upon the back or rump. On the following day the swelling has begun to subside, and soon disappears completely. The exposed skin then hardly differs perceptibly from the adjacent normal skin, except in being slightly smoother, more glistening, and, in certain cases, is a shade darker. Mayer 13 has well described the hardness which is perceptible to palpation as conveying the sensation of a disk or plaque of cardboard inclosed in the skin. This superficial induration persists. After a week or 10 days, a scaly desquamation is observed at the margin of the indurated area, which in the course of the next week becomes sharply defined. The central plaque, composed of dead mummified tissue, becomes sequestrated from the underlying tissue, at first at the margin, eventually over the entire area, until it is finally cast off completely. There is active growth of new hair beneath the slough and one has the impression that the edges are actually pushed up by the growing hairs. The new hairs are usually lighter in color.

(2) After the casting off of the slough, it is discovered that complete healing has taken place beneath the dead tissue. The area is covered with a smooth, glistening epidermis devoid of pigment; and the area remains practically pigment free for at least two months, and perhaps permanently.

(3) Tenderness becomes marked on the second day and seems to last for a long while. Palpation causes a twitching of the skin muscles and evident distress.

e Although an excellent and detailed study of the lesions caused by the application of mustard gas and other vesicant substances to the horses' skin has been carried out by Mayer (report, Extrait du Procès verbale de la séance du Février, 113). it seems desirable to report briefly upon a similar study conducted at Hanlon Field.



(1) Piece excised one hour after exposure.- (a) The epidermis, sweat and sebaceous glands, hair follicles are normal. (b) Corium normal. Blood vessels moderately filled with normal appearing blood cells. Very few wandering cells in the subepidermal connective tissue. Mast cells are fairly numerous about the blood vessels. (c) The only noticeable lesion is a slight edema in the form of a granular coagulum, especially noticeable about the roots of the hair follicles. The edema is somewhat more marked in the subcutaneous connective tissues, where the collagen fibrils are separated by granular coagulated material. Capillaries and lymphatics are not distended; there is no accumulation of leucocytes. No changes are noted in the endothelial cells

(2) Pieces excised six hours after exposure.- The appearances are identical with those seen after one hour, with the following exceptions; (a) The edema is considerably more marked, especially in the subcutaneous tissue. (b) There is a rather marked distension of the lymphatics, which contain a granular coagulum identical with that between the collagen fibrils. There are occasional eosinophiles, mast cells, and mononuclears in the vicinity of the vessels, but no definite inflammatory reaction of a cellular nature.

(3) Pieces excised 11 hours after exposure.- (a) Over the area of maximal edema, in the section, the epidermis is stretched and the papillary folds obliterated. In general there are no abnormal appearances in the superficial epithelial cells, nor in those belonging to the epidermal appendages. A few follicles, however, are definitely injured.

FIG. 234.- Dichlorethylsulphide, 11 hours after exposure. Collagen fibers (C) separated by granular coagulum (oe). Distended lymphatic vessel (L). Emigrated leucocytes (lc)

The earliest recognized alteration is a hydropic change in the epithelial sheaths. Later, there is leucocytic infiltration and destruction of the shaft of the hair. The epithelial cell nuclei become shrunken and pycnotic. The corium in the affected area is somewhat looser in texture than the normal corium and is obviously slightly edematous. The capillaries are moderately filled with red blood cells. There is no excess in the number of leucocytes. The edema of the subcutaneous tissue is well marked, perhaps slightly more intense than in the six-hour specimen. (Fig. 234.) The small vessels show a moderate marginal accumulation of polymorphonuclear leucocytes, with slight emigration into the adjoining cellular tissue. Eosinophiles are quite numerous. A few distended lymphatic vessels are present in the depth of the section.


(4) Piece excised 24 hours after exposure.- (a) The findings do not differ materially from those in the 11-hour section. The edema continues to be marked; there is no increased leucocytic reaction. In fact this feature is less striking than in the 11-hour section. No edema is noted in the corium.

(5) Piece excised 48 hours after exposure.- (a) The edema and distension of the corium has led to the flattening of the overlying epidermis, with partial obliteration of the normal folds. The hair follicles and glands appear less concentrated in the affected area than in the normal skin. The epidermis, itself, is much thinned over the area of the lesion. The pigment is concentrated to such a degree that the finer changes in the cells are obscured. It is possible, however, to make out definite signs of injury to the epidermal cells. This is shown in two ways: A diffuse, intense staining of the nucleus, with loss of chromatin structure, and moderate shrinkage; or there may develop in the cytoplasm an unstained vacuole, which, as it increases in size, causes a concentric compression of the nucleus, and finally leads to its disappearance. In this way the cell comes to contain an unstained, oval clear space surrounded by dense pigment clumps. The early stages of this process, which picks out individual cells, can best be traced at the margin of the lesion or in the less densely pigmented cells of the hair follicles.
(b) It is impossible at this stage of the lesion to distinguish sharply between the living and the dead portions of the epidermis. No clear line of demarcation has been established.
(c) In all the sections examined there appear small areas in which the epidermis is completely destroyed and replaced by a dense collection of leucocytes, the nuclei of which are much fragmented. Pigment granules and the debris of the destroyed cells enter into the formation of these superficial pustules. They may or may not be covered by a thin layer of pigment-containing keratin cells. They appear to originate in the corium, since one may find dense localized accumulations of leucocytes, without destruction or extensive invasion of the overlying epidermis. Usually the pustules are situated adjacent to a hair follicle. One shows a distinct vesicle with a gelatinous thready material, separated from the corium by a dense marginal zone of leucocytes.
(d) Very interesting changes are found in the epidermal appendages. There is partial necrosis of the hair follicles in the affected area. This obviously proceeds from within outwards, the hair shaft and the inner sheath cells often showing marked degenerative changes while the more external cells are still well preserved. So, also, there is in many of the sebaceous glands a well-circumscribed area of necrosis, which regularly affects that portion of the gland which abuts upon the hair shaft, leaving the peripheral cells unaltered. The degenerating sebaceous cells stand out sharply by reason of their brightly red-stained cytoplasm, and the marked shrinkage and subsequent fragmentation of their nuclei.
(e) The sweat glands show less severe changes than the sebaceous glands. Certain acini, however, show necrosis of individual cells with caryorrhectic changes in the nuclei and hydropic swelling of the protoplasm of the cells.
(f) The corium is less edematous than in the earlier stages of the lesion. Here and there can be seen a slight separation of the collagen fibrils, and occasionally a fine fibrinous network can be found under the high power. Distended lymphatic channels containing a pink coagulum can be found in the


deeper layers of the corium. A definite injury to the connective tissue is evident only in the superficial zone immediately underlying the epidermis. Here the nuclei of the connective tissue cells are degenerated, the collagen fibers fused and hyaline, and the elastic fibrils fragmented. Throughout the corium, there is a loose infiltration of leucocytes, chiefly polymorphonuclears. The denser focal accumulations at certain points immediately beneath and within the epidermis have been described. As the emigrated cells approach the surface, their nuclei undergo caryorrhexis.
(g) The capillaries of the corium are well filled with red blood cells. Only a few of the smallest vessels in the papillary layer show fusion and homogenization of the cells suggesting hyaline thrombi. Most of the capillaries even in this zone are quite normal. No definite alterations are found in the endothelial cells.

(6) Pieces excised 72 hours after exposure.- (a) Over the affected area, the epidermis is thinned to about one-third the normal width. Measurements of the normal area show an average width of 17.2 as compared with an average width of 6.2 in the area of the lesion. f
(b) In the normal skin three zones can readily be distinguished, namely: 1. The stratum corneum, consisting of only one or two layers of flat, platelike cells. 2. The stratum granulosum, four to five cell layers deep. 3. The stratum mucosum, composed of one or two layers of cells, with distinct oval nuclei ranged with long axis at right angles to the surface. The distribution of pigment in the normal skin is somewhat irregular, but is in general as follows: While the cells of the stratum mucosum invariably contain abundant pigment, oftenest in the form of a granular crescentic cap at the superficial pole of the nucleus, the pigment is much less abundant in the middle zone. In the flattened horny cells there is often a good deal of pigment in the form of coarser clumps.
(c) The pigment in the affected skin area shows striking differences from this normal distribution. Whether there is an actual increase in the amount of pigment is rather difficult to say. The cells are more closely aggregated and appear compressed, probably by the pressure of the swollen tissue underneath. There is thus undoubtedly a concentration of pigment granules. On the other hand, the pigment granules are clumped into larger masses than the normal granules. The large branching chromatophores in the subjacent tissue are more numerous than beneath the normal area.
(d) The nuclei of the epithelial cells in the thinned-out layer covering the exposed area refuse the stain and have probably lost their structure. Their original outline, however, is still evident from the distribution of the pigment about them.
(e) The epidermis rests upon a wavy structureless pink-staining band, which is not evident in the normal skin.
(f) Immediately beneath this, is a zone of dense cellular infiltration. In it one may distinguish the nuclei of still intact polymorphonuclear leucocytes, but the nuclear material is for the most part fragmented and distorted into wisps and clumps of bizarre shape.
(g) The skin appendages show marked lesions. The sheaths of epidermal cells surrounding the hair follicles are in various stages of disintegration or in
f These numbers refer to divisions on the micrometer eyepieces.


some instances, completely necrotic, and invaded by leucocytes. The sweat glands are also severely damaged, the glandular epithelial cells appearing hydropic, ragged, undergoing caryolysis, and finally disintegrating completely. The sebaceous glands seem to be somewhat more resistant; but individual cells and even entire glands are found showing degenerative changes.
(h) The connective tissue of the corium is severely injured. In the more superficial portions, the nuclei have wholly disappeared; the outlines of the collagen bundles are no longer distinct, the fibers fusing into a homogeneous pink-staining material, loosely invaded by polymorphonuclear leucocytes. Deeper down, the connective tissue is not necrotic; the connective tissue nuclei and the collagen fibrils stain normally; there is, however, considerable edema. The lymph vessels show moderate distension.
(i) Capillaries filled with normal-appearing red blood cells can be traced up to the papillary layer of the corium and appear to be lined with normal endothelium. Whether the small terminal loops are thrombosed cannot be made out clearly because of the dense cellular infiltration; there is no evidence of thrombosis or other vascular injury in the deeper tissues.

(7) Piece removed at autopsy 98 hours after exposure.- (a) The section includes very little normal skin. As the area of the lesion is approached there are numerous small pustular

FIG. 235.- Dichlorethylsulphide, 98 hours after exposure. Superficial pustules (p) at margin of lesion. Increased pigmentation (pg) in adjacent epithelium. Leucocytic infiltration of corium

foci, some wholly within the epidermis itself, others immediately beneath it or involving only the basal portion of the epidermis.(Fig. 235.) These pustules are apparently free from bacteria (Gram-Weigert-safranine). Over the entire area of the lesion the epidermis is thinned and the nuclei stained diffusely and indistinctly. There is again an apparent increase or concentration of pigment.
(b) The superficial zone of corium is very densely infiltrated with leucocytes, although these are separated from the overlying epidermis in most places by a narrow hyaline zone which is free from cells. (Fig. 236.) The nuclei of the leucocytes show extreme caryorrhexis, and are often drawn out into the most irregular shapes, like molten lead. In this mass of broken up nuclear material the connective tissue cells can not be distinguished.


(c) The edema of the deeperportions of theconnective tissueisnotmarked. Degenerative changes in the follicles, sebaceous and sudoriparous glands are evident throughout the section, although individual structures vary in degree to which they are affected. There is a scattered infiltration of polymorphonuclear leucocytes throughout the depths of the corium, with dense localized accumulations about some of the follicles and glands.
(d) The immunity of the blood vessels is again quite striking. Even in some of maximal leucocytic infiltration, capillaries containing normal red-blood cells and showing no distinct lesions, are found.
(e) Reparative changes are not yet present, nor has definite sequestration of the injured tissue begun.

(8) Piece excised seven days after exposure.- (a) Epidermis: Proceeding from the healthy margin, the epithelium as it approaches the area of exposure becomes thinned out, and at the junction, which is rather abrupt, there is a partial exfoliation of the superficial layers. Proceeding toward the center, the epidermis becomes elevated from the underlying corium by an accumulation of leucocytes, mixed with fibrin and cell detritus, immediately beneath it. The leucocytes at the base of the pustule are still preserved; near the surface they are much fragmented and distorted. This area probably corresponds to the indurated ridge which can be felt at this stage in the living animal at the margin of the affected area.

FIG. 236.-Dichlorethylsulphide, 98 hours after exposure. Thinning of epithelium (E), infiltration of corium (c) with leucocytes showing caryorrhexis. Degenerate follicles (H)

(b) Over the summit of the leucocytic accumulation the epidermis is extremely thin, and in places difficult to recognize as a continuous layer being obscured by the mass of nuclear debris. In general, however, the alignment of the epidermal cells is preserved, even when the nuclear staining is lost.
(c) Proceeding centrallybeyond the limits of this pustular ridge, the epithelium is found to be preserved as a thin continuous layer, in which the nuclei are represented by unstained oval spaces, surrounded by pigment. The pigment clumps are coarser and apparently more abundant than in the normal epithelium at the margin. From the basal layer of the healthy epithelium at the edge of the lesion there may be traced under the pustular area and for some distance beyond a tongue of epithelial cells. (Fig. 237.) These form a continuous sheet of irregularly flattened cells with large pale nuclei and fibriliated cytoplasm. They contain little or no pigment. They are continuous with the cells of the hair follicle sheath lying at the same depth, which also are characterized by very large pale nuclei. There appears to be, indeed, a very active growth of new epithelium about the base of these follicles. Mitoses are quite frequent, and the cells not only have hypertrophic atypical nuclei,


but they do not show the orderly alignment and progressive keratinization of the inner layers seen in the normal follicles. Similar changes are noted in the follicles which underlie the central portion of the lesion.
(d) The intense leucocytic infiltration and consequent destruction of the corium at the edge of the lesion has been described. Farther centrally the epidermis rests upon a layer of dense but apparently necrotic tissue in which there is also toward the surface a dense collection of material showing caryorrhexis. The nuclear masses which are derived from invading leucocytes lie beneath a somewhat swollen membrana propria. There is very little infiltration into the cells of the epidermis itself. The necrosis of the superficial zone of the corium is evidenced by the partial fusion of the collagen fibrils into a uniform pink-staining material and by the loss of the connective tissue nuclei.
(e) In this zone of superficial necrosis all the normal skin structures are more or less completely disintegrated. The hair follicles are surrounded by epidermal cells, the nuclei of which are distorted and pycnotic. Often they are invaded by leucocytes, which in turn have become fragmented. The sebaceous glands also show degeneration of the nuclei of the gland cells and are often the seat of dense leucocytic invasion.
(f) The small nerves, so far as can be judged in hematoxylon-eosin and Van Gieson preparations, are severely injured. This is shown by swelling of the endoneurium and perineurium, and by degenerative changes in the sheath nuclei.

FIG.237- Dichlorethylsulphide seven days after exposure. Beginnings equestration of necrotic epithelium (El) and superficial corium (C) by ingrowth of regenerating epithelium (E2). Proliteration of cells of sheaths of hair follicles (H1, H2)

(g) The small capillaries in the necrotic zone are filled with a brownish staining material which, as the healthy tissue beneath is approached, can be seen to be derived from fused and altered red blood cells. This change is limited to the superficial zone immediately beneath the necrotic epidermis; the blood vessels and their contents in the deeper portions of the corium are quite unaltered, even in places where the skin appendages show the effects of the injury.
(h) The elastic fibers in the zone of superficial injury have largely disappeared, only faintly staining fragments persisting. About the regenerating hair follicles, the very characteristic arrangement of the elastic fibers, in the form of parallel bands, joined at the epithelial margin by vertical stays, is no longer recognizable.

(9) Piece excised 14 days after exposure.- (a) The appearances do not differ materially from those in the one week's specimen. Regenerative changes are somewhat more advanced. Epidermis: Over the neighboring healthy


region, there is perhaps a somewhat increased pigmentation. This ceases abruptly as the lesion is approached, and for a short distance near the margin the epidermis is virtually pigment-free. There is also in this region slight hyperkeratosis. (Fig. 238.) The tonguelike continuation of the epidermis can now be traced for a considerable distance beneath the superficial sequestrating dead epidermis. (Fig. 239.) In places, in a study of a number of sections, small islands of flattened epidermal cells which appear to take their origin about partially degenerated hair follicles are found, and seem to be independent or the growing epithelial processes arising at the margin of the lesion. Individual cells of the new-forming epidermis are highly atypical. Hypertrophic cells with nuclei three or four times the normal volume are common. (Fig. 240.)
(b) The original layer of dead, thinned, and deeply pigmented epidermis is still present, attached to a layer of necrotic corium, filled with distorted chromatin masses derived from invading leucocytes and from the degenerated fixed elements of the necrosed tissue. There are many large branching chromatophores. The blood vessels are filled with brown material derived from altered red corpuscles, but their outlines are still preserved.
(c) The demarcation of the dead tissue from the underlying living corium is very sharp and distinct; in many places it is marked by a zone of leucocytic infiltration. In this transitional zone, the collagen fibers are swollen and hyaline. Between them, the connective tissue cells have taken on the character of fibroblasts, the nuclei becoming large and vesicular. Isolated mononuclear and polymorphonuclear leucocytes are scattered amongst the connective-tissue

FIG. 238.- Dichlorethylsulphide, 14 days after exposure. The sequestration of the dead epidermis (E) and the underlying necrotic tissue (C) by in-growth of regenerating epithelium (E:). Desquamated keratin (K)

(d) The blood vessels in places still contain an excess of polynuclears and are frequently surrounded by small cells. There are occasional collections of eosinophiles.
(e) About many of the necrotic hair follicles, there is occurring an active proliferation of epithelial cells, frequently showing mitosis, and often very atypical in size and staining. Regenerative changes are not very evident in the sudoriparous glands, although the secreting cells occasionally show alterations, irregularity of arrangement, darker staining of cytoplasm, hypertrophy of the nuclei of individual cells, which suggest a more active growth. No mitoses were found. There is, however, an active prolifieration of fibroblasts about groups of glands, frequently with collections of eosinophiles and scattered lymphoid cells.


(10) Piece excised 32 days after exposure (obtained at autopsy).- (a) The block includes a loose, button-shaped cap of necrotic issue, 2 to 3 mm. thick which is already detached from the underlying living tissue. Because of the difficulty of impregnating and sectioning this dried, crumbly, material, its microscopic structure could not be studied.
(b) The original site of application is completely covered with new-formed epithelium. This differs from the healthy epithelium at the margin in being distinctly thicker; roughly, from two tolthree times the average width of the contiguous epidermis. The alignment of the cells is irregular. Pigment formation is beginning, but the amount formed is insignificant. The new formed epidermis sends very irregular processes into the corium, whereas in the normal epidermis, the basal boundary forms an even line. The hair follicles are completely regenerated throughout the affected area and do not differ obviously from the normal follicles. The sweat and sebaceous glands likewise show no evidence of previous injury. The superficial layer of corium differs from the normal in its looser texture and in the slightly irregular arrangement of the collagen fibrils, which in the normal corium tend to run parallel to the surface. There are still occasional wandering cells present.


The history of the dichlorethylsulphide lesion, therefore, is somewhat as follows:
(1) The primary visible evidence of injury is an edema affecting the deeper layers of the corium, appearing early, and gradually subsiding. The reabsorption is accompanied by distention of the lymphatic vessels. The exudate is for the most part non- fibrinous

FIG. 239- Dichlorethylsulphide, 14 days after exposure. Same designations as Figure 228

(2) The necrosis of the epidermis takes place very gradually by a process of mummification, in which there is no dislocation of the individual cells. Even after several weeks the basal cells are not disintegrated and their original nuclei are still recognizable as oval clear spaces. There occurs a clumping and concentration of pigment, and possibly during the earliest stages, an increased pigment production.

(3) The necrosis of the epidermis and of the superficial portion of the corium does not appear to depend upon a primary vascular thrombosis. In the early phases of the process, capillaries with intact red blood cells and endothelium are found to extend almost to the surface epithelium. Later they share the fate of the superficial tissues, the blood cells fusing into a brownish


mass, and the nuclei of the endothelial cells shrinking and becoming pycnotic. The deeper vessels, aside from the changes incidental to the inflammatory reaction, show no evidence of injury.

(4) The alterations in the nuclei of the emigrated leucocytes and also of the fixed connective tissue cells in the zone of the injury are very striking.

(5) The penetration of the toxic agent along the shafts of the hair follicles is rendered probable from the fact that the injury to these structures and to the sebaceous glands evidently proceeds from within outwards.

(6) Reparative processes begin toward the end of the first week by an intrusion of a flat tongue of regenerating epithelium between the superficial zone of necrosis and the underlying healthy tissue. The sheaths of the hair follicles play an active role in the regeneration, the cells proliferating and establishing connections with the epithelial processes growing in from the healthy margin. The new epidermis is at first pigment-free and contains many atypical cells. The preservation of the basal portion of many of the hair follicles explains the renewed growth of hair beneath the dead sequestrum.



(1) Twenty-five one-hundredths cubic centimeter of a 10 percent alcoholic solution were applied over areas of shaved skin measuring approximately 0.05 sq. cm. For comparison, the same amount of a 10 percent alcoholic solution of dichlorethylsulphide was applied to a neighboring area. Pieces were excised for histological study after 5 hours, 24 hours, and 48 hours. No study has been made of the late lesions

FIG. 240.- Dichlorethylsulphide, 11 days after exposure. Hypertrophic cells at margin of hair follicles


(1) After five hours the lesion caused by Lewisite involved approximately three times the area of the lesion caused by dichlorethylsulphide and was also more elevated. During excision it was observed that the Lewisite lesions occu- pied roughly three times the area of the mustard-gas burn; the tenderness and induration were more marked, and the pigmentation distinctly deeper. The same differences in intensity were noted after 48 hours, but subsequently became less marked.


(1) Piece excised five hours after application of Lewisite.- (a) The epidermis is distinctly thinned. There is increased pigment, especially at the margin of the lesion, and in certain hair follicles. No degenerative changes in the epidermal


cells, but some of the epithelial cells of the hair follicles show beginning vacuole formation. The most striking feature is a very marked edema of the cutis and as much of the underlying connective tissue as is included in the section. There is wide separation of the connective-tissue fibers and of the epidermal appendages by a granular coagulum. The more superficial capillaries are crowded with normal appearing red blood cells and contain no excess of leucocytes. There is no obvious injury to the endothelium. The deeper vessels, both the small arterioles and the capillaries, show marked changes. They contain red blood cells of altered staining reaction, and are surrounded by loose collections of polymorphonuclear leucocytes. The nuclei of the emigrated cells are pycnotic and hazy. No dilated lymphatics are found in the sections.

(2) Piece excised five hours after application of dichlorethylsulphide.- (a) There are no marked changes in the epidermal cells of the skin or hair follicles. Occasional cells show the presence of an unstained vacuole, with compression of the nucleus. There is very moderate edema of the corium. There is no marked congestion. About a few of the capillaries are loose aggregations of wandering cells, amongst them many eosinophiles. There are no lesions of the sweat glands or sebaceous glands.

FIG. 241.- Lewisite No. 1, 24 hours after exposure. Edema leucocytes in corium (L), congestion of superficial capillaries (B) early degenerative hours after application and thinning of epidermis

(3) Piece excised 24 hours after application of Lewisite (fig 241).- Epidermis: Thinned over lesions, with loss of papillary folds. Alteration of staining in certain areas, increased affinity for eosin. Shrinkage and diffuse staining of nuclei. Vacuolar degeneration of nuclei of individual cells at margin of lesion.
(b) Corium: Papillary edema, with foamy, clear spaces immediatelybeneath epidermis. Leucocytic infiltration, quite dense in places, in all layers of coriuin and in subcutaneous tissue. The predominant cells are polymorphonuclears, eosinophiles are scarce. There is not the extreme fragmentation of nuclei seen in dichlorethylsulphide preparations. (c) Appendages: Hair follicles, same changes as in epidermis. Sweat glands, individual acini show complete necrosis with pycnosis, fragmentation, and eventual complete disappearance of nuclei. (d) Blood vessels: Marked congestion throughout. Platelet thrombi in some of the vessels. No marked distention of lymphatics.

(4) Piece excised 24 hours after application of dichlorethylsulphide.- (a) Epidermis: Vacuolar degeneration of individual cells. Slight flattening. No change in staining reaction. Similar findings in cells of hair follicles. Sweat and sebaceous glands show no lesions. (b) Corium. Moderate edema. Loose and scattered infiltration of wandering cells. Many eosinophiles. (c) Blood vessels: Normal. No marked congestion.


(5) Piece excised 48 hours after application of Lewisite (fig. 242) .- (a) Epidermis: At the edge of the lesion, there is a very abrupt transition between the living epidermis and the dead epidermis. The latter is thinned and stains diffusely and intensely with eosin. Only the nuclear outlines persist, but all chromatin staining is lost. The healthy epithelium has already begun to push its way for a short distance beneath the dead epidermis. There is the same apparent increase in the density of pigment that was noted in the mustard lesions. (b) Epidermal appendages: In the affected area, the epithelial cells surrounding the hair follicles stain. diffusely with eosin. The nuclei are shrunken and diffusely stained. The sebaceous glands are similarly affected. (c) Corium: The superficial portion is moderately edematous; the outlines of the collagen fibrils are indistinct. There is a sparsity of nuclei, and such nuclei as are still present show pycnosis. Leucocytic infiltration is present only at the margin; over the summit of the lesion the corium is practically free of wandering cells. The deeper layers of the corium, on the other hand, are diffusely invaded with leucocytes. These do not show the caryorrhexis which is so striking in the mustard gas lesions. (d) The subcutaneous tissue is very edematous, the fibers being separated by a fibrinous exudate. The vessels are filled with leucocytes and the perivascular sheaths are infiltrated with them. A number of the large veins contain thrombi.

No histological study of the reparative stages of the Lewisite lesions has been made.

FIG. 242.- Lewisite No. 1, 48 hours after exposure. Transition between living epidermis (El) and dead epidermis (E2). Leucocytic infiltration at margin of lesion (Lc), edema of deeper layers (oe)


A comparison of the earlier changes brought about by the two substances in equal concentration shows the following histological differences: 1. Lewisite produces an earlier necrosis of the epidermis. 2. The edema produced by Lewisite is more extensive than that caused by mustard gas, and the exudate is definitely fibrinous. 3. The Lewisite does not so definitely penetrate along the shafts of the hair follicles. 4. The inflammatory reaction appears sooner with Lewisite than with mustard gas and is more intense and deeper. The emigrated leucocytes do not undergo early caryorrhexis. 5. Vascular thrombosis is a more conspicuous feature with Lewisite than with mustard gas. 6. Beginning reparative changes in the epithelium were observed after 48 hours in the case of the Lewisite lesion; they did not begin until the end of the first week after application of corresponding concentration of mustard gas.



The sample of dimethyltrithiocarbonate used for the following tests was a pure specimen, free from dimethylsulphate, furnished by the gas service laboratories.

The concentrations in the test chamber were obtained either by heating the liquid on an electric hot plate or by spraying an alcoholic solution. The atmosphere within the chamber was kept constantly in motion by means of an electric fan. The temperature in the chamber during exposure was about 15° C. The figures for concentration in the table below are nominal: no successful analyses of the air within the chamber were carried out. The boiling point of dimethyltrithiocarbonate is said to be almost exactly the same as that of dichlorethylsulphide and it has been assumed that the vapor pressures of the two substances at various temperatures are of the same order. If experience obtained with dichlorethylsulphide applies to dimethyltrithiocarbonate, one should expect actual concentrations to be about one-third the nominal.

The only animals available were rabbits, guinea pigs, and white mice. The stock of guinea pigs and mice appeared to be quite sound; the stock of rabbits was not so sound as could have been wished; at the time of the experiments, many showed slight nasal discharge and a few died showing signs of bronchopneumonia. The surviving animals were killed with hydrocyanic acid in from two to six days after being exposed. In experiments W-6 to W-10, inclusive, approximately half the total number of animals surviving were killed in two days and the remainder at the end of four days.

TABLE 96.- Effect of dimethyltrithiocarbonate upon animals


In the experiment (W-5) in which a concentration of 1 to 2,850 was used, the symptoms exhibited by the rabbits were: Sneezing and salivation, together with some listlessness and gasping movements which persisted for some time. After two days both of the rabbits had a thick nasal discharge. There was a loss of appetite. The guinea pigs showed lacrymation and sneezing in the chamber, but behaved normally thereafter. The mice breathed deeply while exposed, but on removal from the chamber, although lively, seemed to hold their eyes closed. Later one mouse was found moribund and died.


With lower concentrations the animals might become excited, sneeze, rub their noses, become quiet, and seemed depressed, but appeared perfectly normal after removal from the gassing chamber.


On the basis of the results, tabulated above, it may be concluded that dimethyltrithiocarbonate is a substance of low toxicity. In view of the pathological findings, given below in detail, further conclusions can not be drawn at this time.

Autopsies and histological examinations of the lungs and trachea of 25 rabbits, 35 guinea pigs, and 26 mice used in physiological experiment W are summarized in tabular form as follows:

TABLE 97.- Pathological report


TABLE 97.- Pathological report-Contintued


1. In all experiments with the varying concentrations tried, a large proportion of both rabbits and guinea pigs showed pulmonary lesions, usually a bronchopneumonia of greater or lesser extent, but in a few cases simple edema. The remaining animals were normal grossly and histologically.

2. It is doubtful whether the respiratory lesions observed can be attributed to the inhalation of the dimethyltrithiocarbonate, since: (1) A number of stock rabbits, dying spontaneously about the time the experiments were being conducted, showed respiratory infections with lesions similar to those in the experimental animals.
(2) Two guinea pigs killed with hydrocyanic acid as controls, showed bronchopneumonia lesions of moderate severity, similar to those in many of the exposed animals.
(3) Only 4 mice out of 26 exposed showed pulmonary lesions, and these were trifling. Such a relative species immunity in the case of mice would be highly exceptional.

3. It can not be definitely stated, on the other hand, that the exposure to the toxic gas did not predispose to subsequent infection. This possibility is borne out by the fact that in a small proportion of the cases edema, with little or no inflammatory exudation into the alveoli, was found a type of lesion difficult to attribute to spontaneous infection.

4. It would be desirable to carry on further experiments, using sounder stock and carrying the observation over a longer period.



A telegraphic request was received to apply tests on the French species of the common European snails and slugs, since recent laboratory reports indicated that such animals were very sensitive to dichlorethylsulphide, reacting to it in a concentration of 2 parts per million.

Laboratory facilities were not at hand at the time the request was made, so a preliminary study was made under circumstances somewhat similar to field conditions.

The specimens employed were the common red, forest slug (Arion rufus) and the common white, edible type of snail (Helix asperosa). The experiment was conducted as follows: (a) Direct application of the crude mustard oil from a German shell to an area on the dorsal and ventral surfaces of 3 slugs and 3 snails. (b) Exposure of 12 slugs and 12 snails to the vapor of the same substance. (c) Notes on these animals with regard to such common factors as smoke, concussion, rapid changes in temperature and lights, as well as the character of the materials upon which they may be placed.


The red slug was touched with a droplet of crude dichlorethylsulphide on the dorsal and ventral surfaces near the cephalic end. Immediately there began a slow general contraction of the entire body and an abundant secretion of transparent, viscid fluid which dried on the fingers of the examiner very much as does collodion. This secretion very quickly formed a large drop at the point where the dichlorethylsulphide was applied. Aside from this local feature there did not appear, either early or late, any local change such as a marked change in color, edema, ulceration, or exfoliation. The slugs were all somewhat shrunken in size, and a general change in color from a yellowish-red to a brownish-red took place after several hours had passed. This may have been due simply to the contracted state of the body, for at a later observation they approached their normal color. At the end of two weeks there was nothing more to be noted.

The white snail was treated in the same way as the red slug. Upon application of the dichlorethylsulphide, it immediately drew its body well back into its shell, leaving a ball of transparent, foamy, viscid material closing the entrance. It was easily made to retract and disappear in its shell by simple irritation with a straw or other foreign body, but the secretion mentioned was rather remarkable. Evening and daytime observations during the next three days failed to find any of them protruding from the shell, but thereafter they appeared in the evenings or when placed in a shaded, moist, grassy spot. They did not make as long an excursion as usual, however. At the end of three weeks all were still alive and revealed only a slight brownish-yellow pigmentation or the ventral surface.



A tight wooden box of about one-quarter cubic meter capacity was employed. It was closed with a roof made of gelatin-filled cordon wire, and over the grass-covered ground floor was fixed a screen wire (not filled with gelatin) which rested a few centimeters above the ground. Over this were scattered some fresh twigs and leaves. Before adjusting this floor and admitting the animals about 15 c.c. of crude dichlorethylsulphide was scat- tered over the ground. In this way the snails were exposed to the vapor of dichlorethylsulphide under somewhat modified field conditions. The red slugs showed more or less activity during the first few hours, but the following day most of them were quiet and hidden away in the leaves, even late in the evening. Two were found in direct contact with the wire screen floor, and these were shrunken in size and of a reddish-brown color. These slugs, when released at the end of three days, were able to crawl away except for the pair found lying directly on the screen floor. The common white edible snails, when tested under the same conditions, rather slowly withdrew into their shells, leaving a foamy, viscid, transparent ball of fluid covering the shell hole. The following morning a few of those among the leaves had reappeared, and the exposed portions of body margins, as well as the ventral surface, were discolored a yellowish-brown and covered with a turbid film. At the end of two days they were removed and placed on a moist block of wood covered with a screen cage that would permit short excursions over grass about the block. During the three weeks these snails were under observation only two or three left the block, and very few appeared from the shell in so far as evening observations were concerned. By touching them with a straw they would quickly recede to a greater depth in the shell. Between 10 days and 3 weeks afterwards all had died. The ventral surface, margins, and scattered local areas were dry, brown, and hard.


It was found during the course of this study that these animals reacted easily to many factors. Their movements when exposed to dichlorethylsulphide were quite like those due to common irritation such as exposure to tobacco smoke, sudden extreme changes in temperature, and light, jarring, strong winds, etc. It would appear unsafe, therefore, to place too much reliance on their immediate behavior when placed in the presence of field concentrations of dichlorethylsulphide.


The red forest slug (Arion rufus) is much more resistant to dichlorethylsulphide action than the snail (Helix asperosa). The above experience does not indicate that these snails and slugs can be made use of as field indicators for dichlorethvlsulphide.


The suits used in these experiments were made of two layers of "canaburg," a loosely woven cotton fabric, the outer layer of which was impregnated with simplexene B (45 parts rosin, 55 parts rosin oil). 14 The prepared fabric was quite freely permeable to air, and the suit did not interfere with heat loss from the body to the same extent as do the impermeable suits hitherto used.


It was not anticipated by the designers that the fabric would protect against liquid dichlorethylsulphide; consequently the first experiments dealt solely with protection against vapor. Before submitting the suits to extensive trial at the front request was made for data concerning the possibility of protection against the liquid, and certain of the experiments herein recorded dealt with the possibility of protection against liquid and spray.

The conclusions which the experiments appear to justify follow:



The horse's skin was completely protected by suit fabric from 15 minutes' exposure to saturated dichlorethlysulphide vapor at about 13° C. Protection was not complete against 25 and 30 minutes, exposure at 11° C. Wet suit fabric was somewhat less protective than dry. (Experiments 1-5.)


By the static method of testing, human skin was completely protected against 20 minutes' exposure to air saturated with dichlorethylsulphide vapor. Protection was almost complete against 40 minutes' exposure. By another method of testing, complete protection against an hour's exposure to a high concentration was obtained. (Experiments 6-9.)


Partial protection was afforded to persons wearing the protective suit in nominal concentrations of 1 to 100,000, 1 hour, and 1 to 31,000, 30 minutes. Protection was least adequate for the scalp, scrotum, and inner surface of thighs. (Experiments 10-13.)


More dichlorethylsulphide was absorbed by the protective fabric than by khaki cloth when both were suspended in air saturated with dichlorethylsulphide vapor. (Experiments 14 and 15.)


One layer of suit fabric, the outer surface of which was wet with yellow cross shell filling, might remain in contact with the horse's skin for between 16 and 30 minutes without causing lesions. (Experiment 16.)

When the suit fabric, in direct contact with human skin, was sprayed with yellow cross shell filling, it might remain in contact for 10 minutes in some cases without producing lesions. In other cases erythema was produced by contact of this duration. Severe reaction was apt to be produced by contact of 2 to 30 minutes.

When the suit fabric, placed over the other clothing (blouse, shirt, and undershirt), was similarly sprayed, it might remain in contact for 30 minutes in some cases without producing reaction. Experiment 20 indicates that sensitive persons would be injured in such a test. (Experiments 18-22.)

One subject, fairly insensitive to diehlorethylsulphide was completely protected during 50 minutes in which he was enveloped for from 30 to 60 seconds


  in the visible clouds produced by separate explosions of six 75-mm. dichlorethylsulphide shells. Another subject, decidedly more sensitive to dichlorethylsulphide, developed mild erythema as the result of the same exposure. (Experiments 23 and 24.)

When 600 c.c. of yellow cross shell filling were dispersed in a manner closely similar to that in which it was dispersed when a German 77-mm. yellow cross shell was exploded at rest, men wearing the protective suits, together with boots, gloves, helmets, and respirators, might stand 4 to 6 paces away from the explosion, receive liberal spattering of spray and droplets, wear the suits for a further period of from 20 to 60 minutes, and not be injured seriously. This conclusion was based on three experiments in which eight persons were exposed. The length of time during which a person might safely wear the suit after such exposure obviously depended on the quantity of shell contents with which he was spattered, and judgment was necessary in applying this conclusion to comparable field exposures. (Experiments 25-27.)

Men wearing the protective suit, together with boots, gloves, helmets, and respirators, might walk through woods which had been heavily shelled with yellow cross for from 20 to 30 minutes, and might keep the suits on while working for another 30 minutes with a fair chance of escaping injury. If too great confidence were placed in the protective power of the suit serious casualty might result. (Experiments 28-29.)

After being contaminated by shell burst or by contact with undergrowth in woods shelled with yellow cross, the protective suit might retain dangerous amounts of dichlorethylsulphide for as long as 48 hours and doubtless longer. It would be unwise to wear it as long as it gave off the odor of dichlorethylsulphide. (Experiments 30-31.)


On the basis of these conclusions it was thought that the protection afforded by this suit might enable men to work for several hours without serious danger on open ground heavily contaminated with dichlorethylsulphide; to walk for half an hour through thick woods which had been heavily shelled with yellow cross, or for a longer time if the woods were open or if they had been lightly shelled. If a man wearing the suit should be heavily spattered with dichlorethylsulphide from a shell burst, or if the fabric became actually wet with dichlorethylsulphide from any cause; the suit should be removed as soon as possible.




Air saturated with vapor of dichlorethylsulphide at outdoor temperature was applied to the shaved skin of a horse. A layer of the suit fabric was interposed between the mouth of the bottle containing cotton wet with dichlorethylsulphide and the skin. In some control tests, khaki cloth was used instead of the protective cloth; in other controls the vapor was applied directly to the skin.


EXPERIMENT 1.-Vapor from pure dichlorethlysulphide, and from yellow cross shell fillings; temperature, 13.5°C.; exposure, two hours. Skin protected by one layer of suit fabric. Result.- Severe reaction began to develop in both tests 15 minutes after exposure.
  EXPERIMENT 2.- Vapor from pure dichlorethylsulphide; temperature, 9°C.; exposure 30 minutes. (a) Skill protected by one layer of suit fabric. (b) Skin protected by one layer of khaki. (c) Skin unprotected. Result.-The reaction which developed in (a) was distinct but decidedly less than that in (b) or  (c)
EXPERIMENT 3.-Vapor from pure dichlorethylsulphide; temperature, 12.8°C.; exposure 15 minutes. (a) Skin protected by suit fabric.( b) Skin protected by khaki. (c) Skin unprotected. Result.-No reaction in (a); definite reactions in (b) and (c).
  EXPERIMENT 4.-Vapor from pure dichlorethylsulphide; temperature, 11°C. (a) Skin protected by suit fabric; exposure, 25 minutes. (b) Skin unprotected; exposure, 5 minutes. Result.-Reaction in (b) slightly greater than in (a). The reactions were of the same order.
EXPERIMENT 5.-Vapor from pure dichlorethlysulphide; temperature, 11° C.; exposure, 15 minutes. (a) Skin protected with suit fabric which was thoroughly wet with water. (b) Skin protected with khaki. (c) Skin unprotected. Result.-Slight reaction developed in (a); definite reactions developed in (b) and (c).


EXPERIMENT 6.- Subject R. Exposure of forearm to saturated vapor at 9°C. Static method. (a) Skin protected with suit fabric; exposure, 20 minutes. (b) Skin protected with khaki; exposure, 10 minutes. (c) Skin unprotected; exposure, 10 minutes. Result.-No reaction in (a). Erythema in (b) and (c).
  EXPERIMENT 7.- Subject G. Saturated vapor at 90 C. Static method. (a) Skin protected with suit fabric; exposure, 40 minutes. (b) Skin protected with khaki; exposure, 15 minutes. (c) Skin unprotected; exposure, 10 minutes. Result.-Very slight erythema in (a). Marked erythema in (b) and (c).
EXPERIMENT 8.- Subject G. The whole forearm was protected by two layers of the suit fabric except at one spot 112 inches in diameter where one layer was cut away. For 30 minutes it was held in a large tin box in which an excess of liquid dichlorethylsulphide had been placed and warmed with an alcohol lamp. Result.-No reaction developed.
EXPERIMENT 9.-Subject R. Forearm, protected as described under experiment 8, was held for one hour in tin box containing dichlorethylsulphide vapor. Liquid dichlorethylsulphide had been standing in a shallow dish within the box for 16 hours. Result.- No reaction developed.


(a) Concentrations of dichlorethylsulphide were raised in a chamber of 19.7 cubic meters capacity. The walls of the chamber were covered with "beaver board" and doubtless absorbed a portion of the vapor. The actual concentrations were probably materially lower than calculated. (b) Subjects exposed in the chamber wore the protective suit over all their other clothing except the blouse, which was not worn. The shirt sleeves of one arm were rolled above the elbow and a small circular hole was cut through the protective sleeve, so that a circle of unprotected skin was exposed.
(c) Protective boots, gloves, and the small box respirator were also worn.
(d) Concentrations were raised by heating pure dichlorethlysulphide in a casserole by an alcohol lamp. The subjects were in the chamber during the 19 minutes or more necessary for evaporating the dichlorethylsulphide, but recorded times of exposure refer to time after the concentrations were raised.
(e) Two rabbits were exposed in the chamber during each experiment.

EXPERIMENT 10.- Subjects R and G. Nominal concentration of dichlorethylsulphide, 1 to 100,000, raised by heating 1.1 c. c. purified dichlorethylsulphide. Time of exposure, 1 hour. Result.-No unmistakable erythema of exposed spot on right forearms. Distinct


erythema appeared about old dichlorethylsulphide scars on the protected left arm of one subject. Rather severe itching of old dichlorethylsulphide scars was noted by both subjects during the evening following exposure. One rabbit showed severe conjunctivitis the following morning; the other showed little.
EXPERIMENT 11.- Subjects R and G. Nominal concentration 1 to 31,000, raised by heating 3.5 c. c. purified dichlorethylsulphide. Time of exposure, 30 minutes. R wore same suit as in experiment 10; G wore a new suit. Results.-Very faint erythema of exposed spot on right arm in both subjects. Annoying irritation of old dichlorethylsulphide scars on left arm, of scalp, and of scrotum was experienced during evening following exposure. G developed slight erythema of forehead above line of face mask of respirator. Both rabbits showed severe conjunctivitis 18 hours after exposure; 24 hours later excessive nasal discharge was noted.
EXPERIMENT 12.- Subjects R and D. Subject D is known to be more sensitive to dichlorethylsulphide than either R or G. Nominal concentration, 1 to 31,000. Exposure, 30 minutes. Unprotected areas on right arm larger than in previous experiments. R wore same suit as in two previous experiments. D wore G's suit of experiment 10. Result.- Subject D: Seven hours after exposure severe itching of scalp, body, and thighs began. Areas about old dichlorethylsulphide scars and scrotum were especially annoying. Twenty hours later erythema of unprotected spot on right arm was distinct. There was slight erythema of scalp, inner surface of thighs, and of scrotum. Irritation was such as to prevent sleep for two nights but no blistering occurred. Subject R: Erythema of exposed spot on right arm. Annoying itching of scalp, scrotum, and inner surface of thighs. Both rabbits developed very severe conjunctivitis and purulent nasal discharge.
EXPERIMENT 13.- Subjects R and K. (K is known to be extremely sensitive to dichlorethylsulphide.) Nominal concentration, 1 to 31,000. Exposure, 30 minutes. Subject R wore a pair of trunks cut from another protective suit in addition to the protective suit itself. Result.-Subject R: Aside from slight irritation of the scalp, protection was complete. Subject K: Erythema of exposed spot on right forearm developed in three hours; it was intense on the following day. Diffuse erythema of right forearm, left wrist, neck, and forehead. The skin of neck became swollen, but no blistering occurred.


EXPERIMENT 14.- Immediately after coming out of the chamber concentration in experiment 12, pieces of cloth were cut out of the sleeve of the suit worn by R and bound to the unprotected skin of the forearm of subject G and worn for one and one-half hours. Result.- No reaction developed.
EXPERIMENT 15.- Pieces of khaki cloth and of protective fabric were suspended in a jar containing air saturated with dichlorethylsulphide vapor for 30 minutes. Small circles of each were cut out and bound on the forearms of two persons and worn for two hours.



EXPERIMENT 16.- The skin was shaved, washed with soap and water and alcohol, and dried. One layer of protective fabric was laid over the shaved area, and over this was placed a layer of impermeable fabric in which five holes had been cut, each 1½inches in diameter. German yellow cross shell filling was sprayed with an atomizer so that the exposed circles of protective fabric were actually wet. The impermeable fabric was removed at once. The wet circles of protective fabric were cut away after contact with the skin for 5, 10, 15, 30, and 60 minutes, respectively. Temperature was 120 C. Result: Contact for 5, 10, and 15 minutes caused no reaction. Contact for 30 minutes caused a slight reaction. Contact for 60 minutes caused a severe early reaction, which receded decidedly during following 24 hours.


EXPERIMENT 17.- Subject R. The forearm, bare save for a layer of suit fabric, covered by impermeable fabric in which were cut four holes, each 1 inch in diameter. German yellow cross shell filling was sprayed by an atomizer directly against the exposed protective fabric. One circle was cut away after contact for 1 minute, the others after 2, 5, and 10 minutes' contact, respectively. Result.-No reaction developed.


EXPERIMENT 18.- Subject R-2. Like experiment 17. Contact with sprayed suit fabric for 10, 20, and 30 minutes. Result.-All three exposures gave moderate erythema, but no blistering. EXPERIMENT 19.-Subject G-2. Like experiment 17. Contact with sprayed suit fabric for 10, 20, and 30 minutes. Result.-Ten minutes' contact caused erythema; 20 and 30 minutes' contact caused erythema followed by blistering, which was severe in the case of the 30-minute exposure.
EXPERIMENT 20.- Subject R (known to be extremely sensitive to dichlorethylsulphide). Suit fabric placed on arm over 0. D. shirt sleeve and woolen undershirt sleeve. Sprayed with yellow cross shell filling as in experiment 17. Contact for 10, 15, and 20 minutes. Result.-Diffuse erythema over whole forearm developed in 5 to 7 hours, which was most severe at points directly under sprayed areas. Erythema spread up the arm and shoulder. The arm became much swollen, and the swelling persisted for about 5 days. A blister developed on the area where contact lasted 20 minutes.
EXPERIMENT 21.- Subject R. Forearm covered by blouse, shirt, and undershirt. Suit fabric, laid over these, sprayed as in experiment 17. (Contact for 10, 20, and 30 minutes.) Result.-No lesions.
EXPERIMENT 22.- Subject W. Like experiment 21. Contact 30 minutes. Result.- Very slight erythema.


EXPERIMENT 23.- Two subjects (C and V) wearing protective suits, boots, gloves, and respirators, ran into the cloud produced by exploding a 75-mm. dischlorethysulphide shell at rest. They walked about the crater for 1 to 2 minutes. The suits were then removed. Result.-No lesions.
EXPERIMENT 24.-Two subjects (R and C), wearing protective suits, boots, gloves) and respirators, ran into the cloud produced by each of six 75-mm. dichlorethylsulphide shells exploded singly at intervals of about 10 minutes. They were in the visible cloud from each shell for from 2 to 10 seconds and walked about the crater of each shell for from 1 to 2 minutes after the cloud had disappeared. The suits were worn for a total of 50 minutes. Result.-Subject R developed no lesions. Subject C: Erythema developed in about 3 hours on arm, trunk, knees, and scrotum. No blisters. Erythema persisted for 48 hours.


EXPERIMENT 25.- The contents of a German 77-mm. yellow cross shell (600 c.c.), contained in a tin can, were dispersed by detonating the picric acid gained from the same shell immersed in the liquid. In this way the explosion of a 77-mm. yellow cross shell was reproduced without danger from shell fragments. Three subjects, R, B, and V-2, wearing protective suits, boots, gloves, helmet, and respirator, stood 6 paces down wind from the "shell." The shell contents were distributed as a cloud of small droplets and fine spray, which enveloped the subjects. Subjects R and V-2 were well sprinkled about left arm, shoulder, and back; only slightly sprinkled below the waist. Subject B was well sprinkled from head to foot. B removed his protective clothing 15 minutes after the burst; V-2, 18 minutes after burst; R, 22 minutes after. Other clothing which had been worn during the test was kept on during the rest of the day. Result.-No lesions in any subject.
EXPERIMENT 26.-The contents of a German 77-mm. yellow cross shell (600 c. c.) were dispersed exactly as in experiment 25. Wind velocity was 7 m. p. s. Three subjects, B-2, C-2, and W, wearing protective clothing, respirators, and helmets, crouched 6 paces down wind from the "shell." Subject B-2 was liberally sprinkled from head to foot. C-2 was well sprinkled over left arm and side. W was well sprinkled about the waist. B-2 wore his protective clothing for 20 minutes, C-2 for 30 minutes, and W for 40 minutes after the burst. Other clothing worn during exposure was kept on during the rest of the day. Result.-Subject B-2 showed slight erythema of back, arm, and scrotum, which began 4 hours after exposure and persisted several days. Irritation was not sufficient to prevent sleep. Subjects C-2 and W developed moderate erythema of back, which persisted for several days, but which caused very little annoyance.

e In experiments 25 to 29 Capt. W. A. Bush, R. E., rendered valuable assistance. He devised and carried out the experiments with shell bursts, the effects of which were studied.


EXPERIMENT 27.- Like experiments 25 and 26. Two subjects, R and V-2, wearing protective clothing, stood 5 paces down wing from "shell" containing 400 c.c. yellow cross shell filling. Another explosion of 300 c.c. of yellow cross shell filling was immediately arranged, the subjects standing 4 paces away. Both subjects were scantily sprinkled. Protective clothing was removed one hour after the first explosion. Result.-Neither subject developed any lesion.


EXPERIMENT 28.- A wooded area, roughly pentagonal in shape and including about 4,000 square feet, was wired off. Except for a small space in the center it was densely covered with small beech saplings, scrub oak, and thorn bushes. It was impossible to walk about in the area without constantly brushing against the twigs and branches. A 150-mm. German howitzer yellow cross shell was exploded in the center of the area. Temperature was 13°C. and wind 2 meters per second. Eighty minutes later, three subjects, R, V-2, and H, wearing protective clothing, boots, gloves, helmets, and respirators, entered the area and walked for 20 minutes, taking pains to go through the thickest undergrowth. On emerging, they removed respirators and helmets and threw back the hood of the suit. The protective suit was worn during a further walk of half an hour and then removed. The temperature during exposure and the subsequent walk was 150 to 16.5 0 C. and the air was very humid; all three became very sweaty. The other clothing worn during the test was kept on during the rest of the day. No precautions were taken other than washing face and hands after removing protective clothing. Result.-No lesions in any subject.
EXPERIMENT 29.- On the day following experiment 28, another 150-mm. German howitzer yellow cross shell was exploded in the same place. Temperature 8.5°C.; no wind; fairly heavy fog. Thirty minutes later, three subjects, W, H, and V-2, wearing protective suits, etc., entered the area. They walked about for one hour, taking pains to rub against the undergrowth and to force their way between the trunks of closely set saplings. One subject (W) stated that whenever he saw a black splash on a tree or bush he rubbed against it. On emerging respirators were removed, but the protective clothing was worn during a further walk of 25 minutes. All became very hot, sweaty, and uncomfortable. On removing protective suits, hands and faces were washed; no other precautions were taken. Result.-Five hours after exposure, subjects W and H showed marked erythema of shoulders, buttocks, thighs, and knees. Subject V-2 showed no lesion. All bathed thoroughly with warm water and soap. Some of the erythematous areas were treated with I per cent potassium permanganate solution. Twenty hours later W has a large blister on right shoulder, a small blister on left wrist, small blisters about the penis, erythema of back, buttocks, thighs, scrotum, and knees. H had large blisters on back and shoulder, small blisters on scrotum, erythema of buttocks, thighs, and knees. Slight erythema of portion of face covered by respirator mask. V-2 had no lesions. W and H were sent to the hospital.


  EXPERIMENT 30.- One of the suits worn in experiment 25 gave distinct odor of dichlorethylsulphide six hours after the experiment. In the meantime it had been hanging in a large airy building. A small piece of cloth was cut from it, bound to arm of subject R, and worn for 4 hours. Result.-Marked erythema had developed before it was removed, and a blister appeared two days later.
  EXPERIMENT 31.- Twenty-four hours after experiment 29, a piece of the fabric from one of the suits worn in that experiment, which still gave distinct odor of dichlorethylsulphide, was bound to the arm of subject D, and worn for an hour. Result.-It produced erythema, but no blister.
Forty-eight hours after the experiment another piece was cut from the same suit, still giving the dichlorethylsulphide odor, bound to the arm of subject R and worn for four hours. Result.-Erythenia developed during the four hours, and severe blistering occurred subsequently.
The suit, during the 48 hours following the experiment, was hanging in a large airy building.



The value ascribed by some observers to sag paste and other ointments as protection against dichlorethylsulphide and the dissenting views held by others, led, in view of the practical importance of the subject, to the making of some independent observations. The ointments tested were: (1) Sag paste (zinc stearate); (2) calcium hypochlorite (10 percent in white petrolatum, U. S. P.); (3) petrolatum (unmedicated). The first named was chosen because it had been selected officially for extensive field tests. Calcium hypochlorito ointment had been found effective experimentally by the French and English. Petrolatum, on the other hand, had been found useless or positively deleterious.

The first experiments were made on the skin of the horse, using alcoholic solutions of dichlorethlysulphide of various concentrations. This method was tried in the hope of obtaining at numerical expression of degree of protec- tion. In later experiments on the horse and on man the static method of exposure to saturated dichlorethylsulphide vapor wtas employed. In the experiments on man, exposures to Lewisite I were included.


Three tests were made to determine the protection afforded by sag paste, calcium hypochlorite ointment, and petrolatum against solutions of dichlorethylsulphide in absolute alcohol. In the method employed, a definite constant amount of dichlorethylsulphide to be evenly applied to a constant known area of the skin with which it was left in direct contact by the evaporation of tho alcohol in which it was dissolved. In the first experiment the efficacy of sag paste was estimated by the application of varying percentages (2 and 3 percent) of dichlorethylsulphide to treated and control areas. With the 1 percent solution no difference was noted between sag paste and control areas: with the 2 percent and more definitely with the 5 percent solution the treatment by sag paste delayed the onset of swelling and, up to 24 hours, retarded its full development. At 48 hours and subsequently the reaction in all areas was practically identical.

In the second experiment sag paste was compared with petrolatum; the former again retarded the onset of reaction to percentages of dichlorethylsulphide above 1 percent when compared with petrolatum, but this favorable effect was not noticed at the end of 24 hours or subsequently.

In a third experiment calcium hypochlorite ointment wtas compared with petrolatum, but no difference in the lesions produced by the dichlorethylsulphide was observable. When, however, a definite percentage of dichlorethylsulphide was added to calcium hypochlorite ointment and petrolatum, respectively, and weighed amounts of these ointments were applied to the skin of the horse, the former caused much less swelling and induration than the latter. This effect, however, wais clearly due to a direct chemical breaking down of the dichlorethylsulphide and not to a true protective action.

Series II.- Since the application of solutions gave such inconclusive results, the effect of dichlorethylsulphide vapor was then tried. Vapor tests have the added advantage of simulating more or less closely conditions likely to be met with in the field. Experiments 4 and 5 were devoted to determining


the most suitable technique, under the circumstances, for applying vapor by the "static" method, and particularly the optimum time of exposure With the temperature at 20°C. or thereabouts an exposure of 5 minutes is sufficient to produce a definite lesion in 24 hours or less, but if the temperature of the air is much below 15°C. it is better to prolong the exposure to 7 or 10 minutes or to make use of a water jacket (at 20°C.). Experiments 6, 7, 8, and 9 were performed successively to determine the protective value of sag paste and petrolatum against saturated dichlorethylsulphide vapor, using untreated skin areas for controls. In experiments 6 and 7 the ointments, irrespective of their composition, seemed to favor a more prompt reaction to the dichlorethylsulphide, while in 8 and 9 even this distinction between the exposed areas was absent. In experiment 9 the ointments were washed off with soap and water five minutes after the exposure to dichlorethylsulphide, but this did not delay or prevent the swelling and induration. While the tabulated (early) results were unfavorable to both ointments, yet subsequently the sag paste areas (experiments 6 and 7) seemed less swollen and indurated than the remaining areas. Other tests (experiments 10 to 14) intended to determine the protective value of hypochlorite ointment, etc., were also negative.

Series III.- Up to this time one horse had been employed. In order to test further the effects of the ointments mentioned and to exclude the influence of idiosyncrasy, the experiments were repeated in two other horses, one with a highly sensitive skin, the other with a tough leathery hide. The results of experiments 15 to 18 were inconstant, though both sag paste and hypochlorite ointment seemed at times to afford slight protection when judged by the lesions at 24 hours. Petrolatum on the other hand seemed to be constantly devoid of protective powers. Further tests of hypochlorite ointment (experiments 19 and 20) were negative.

It was concluded that the protective ointments which were examined are of little or no value in protecting the skin of the horse against solutions of dichlorethylsulphide or against air saturated (20°C. approximately) with dichlorethylsulphide vapor. Whatever value sag paste may have in this connection is manifested either by delaying the onset of symptoms or in slightly modifying their severity.

Experiments in man to test the protective action of sag paste, calcium hypochlorite ointment, and petrolatum against dichlorethylsulphide, and, incidentally, against Lewisite I vapors.

In experiment 29, five individuals with varying susceptibility to dichlorethylsulphide and Lewisite were tested by the vapor or static method, using very small tubes. For the untreated or control areas and the areas treated with petrolatum an exposure of 10 minutes was allowed; for sag paste 15 minutes; for hypochlorite ointment 20 minutes. The intention was to secure distinct lesions in each instance without causing undue vesiculation in the central areas. In estimating the results, due allowance was made for the difference in exposure. Details of technique and tables are given below. From these tests the general conclusion was drawn that sag paste and hypochlorite ointments afforded moderate protection against the vesicant vapors tested which was only absolute when the control lesions were minimal. Petrolatum was found to be either indifferent or positively deleterious. Further tests with this substance were therefore deemed unnecessary, particularly as this finding is in agreement with the observations of others.


In experiments 30 and 31, three individuals tested in each instance, the technique was slightly altered in that larger tubes were employed and the exposures were made uniform and simultaneous. The new sag paste issued by the Purchase and Supply Division of the Chemical Warfare Service was used in these tests. Here again sag paste gave moderate protection against dichlorethylsulphide in nearly three-fourths of the tests and against Lewisite in about one-half of the tests. The results with hypochlorite ointment were similar but less uniform; sometimes the protection was very marked, while in other instances the action of the vesicant was intensified.


(a) In man, therefore, both sag paste and hypochlorite ointment afford a certain degree of protection against the vesicant vapors under consideration; this is seldom or never absolute and is entirely absent in one-third to one-half the cases. Sag paste is preferable, as it is certainly nonirritating and probably more permanent than the hypochlorite ointment.
(b) In testing the relative value of protective substances it is necessary to use human skin. While the skin of the horse is more sensitive, it is probable that minor effects are missed since erythema if present can not ordinarily be distinguished.


(1) Weekly Report No. 13, Section V-1, and No. 14, Section V-1, C-F. On file, Technical Division, Chemical Warfare Service.
(2) Summary Report of Hanlon Field, Chemical Warfare Service, Vol. IV, Field Gas and Artillery Sections. On file. Technical Division, Chemical Warfare Service 02 1.11/ 4
(3) Summary Report of Hanlon Field, Chemical Warfare Service, Vol. 1, Chemistry Sec- tion. On file, Technical Division, Chemical Warfare Service 021.11/ 4
(4) Weekly Report No. 10, Section V-1. On file, Technical Division, Chemical Warfare Service.
(5) Weekly Report No. 12, Section V-2. On file, Technical Division, Chemical Warfare Service.
(6) Weekly Report No. 21. On file, Technical Division, Chemical Warfare Service.
(7) Special Report No. 50. On file, Technical Division, Chemical Warfare Service.
(8) Lynch, V.: American University Experiment Station, Monograph No. 1, Revised, 317.
(9) Winternitz, M. C., Finney, W. P., jr., and Wislocki, G. B: Pathology of Poisoning by Mustard Gas Ph. 113. On file. Technical Division, Chemical Warfare Service.
(10) Muratet and Fauré-Fremiet: Sur les variations numeriques et les modifications morphologiques des elements figures du sang au cours de l'intoxication rapide par l'ypérite. Rapport de M. Mayer.
(11) Jolly: Personal Communication through M. Mayer.
(12) Zunz, E.: Les gazes. Annales et bulletin de la Société des sciences médicales et naturelles de Bruxelles, October, 1919, 66-72.
(13) Mayer: Report-Extrait du Procès verbale de la séance du Février, 1913.
(14) Description of suit designed by Capt. R. G. and 2d Lieut. T. M. Knowland, C. W. S. Research Division, Chemical Warfare Service, XXIII, S-2.