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 XVI




The extensive use of chemical substances in gas warfare led to a systematic study of compounds which irritate the skin. The various compounds submitted by the chemical offense section of the Chemical Warfare Service were compared with dichlorethylsulphide (mustard gas), imitating field conditions in warfare as closely as possible; that is, the skin irritant efficiency of the new compound as compared with dichlorethylsulphide was determined. A brief summary of the methods used and results obtained is here presented. This may be of interest in toxicology and pharmacology, particularly since it is believed that about one-half of the 70 compounds studied have not been previously described.



This was made use of principally in chronic experiments with various animals. The method was devised by Lynch1 at the American University. A 5 sq. cm. area of skin suitably prepared (by shaving and washing) was selected and to this was applied 0.005 c.c. (from a pipette graduated to deliver 0.005 to 0.05 c. c.), or 5 mgm., of the compound, equivalent to the application of 0.001 c. c. (or 1 mgm.) per square centimeter of skin. No precaution was taken to confine the vapors of volatile compounds. The time of appearance of hyperemia and other changes was noted, and the extension and severity of all changes was studied quantitatively as much as possible from day to day and the results expressed in terms of square centimeters of area involved. Dichlorethylsulphide was chosen as the standard by which the irritant properties of other compounds were judged.

On human skin the tests were made by touching a small area with a fine glass rod dipped into the irritant. This method could not be used quantitatively.


Preliminary tests with vapors of volatile compounds were made by two methods devised by Lynch.2 One method consisted of placing a small excess of the compound on a plug of cotton in a test tube of 1 cm. diameter and 5 cm. from the opening of the tube. This was allowed to stand long enough to secure a saturation of the column of atmosphere in the tube and then the tube was applied for variable lengths of time to the skin.

The other method consisted of placing a small excess of the compound on a plug of cotton in the bottom of a test tube (1 c m. diameter and about 10 cm. long) which was inclosed in at water jacket made from it long test tube (20 by 180 cm.). The tube was stoppered and allowed to stand for one hour at about

a These investigations were conducted by the dermatological unit, section of pharmacology and toxicology, Chemical Warfare Service, American University Experiment Station. Washington, D. C. They were reported in The Journal of Pharmacology and Experimental Therapeutics, Baltimore, Md., 1919, xiv, 228.


25° C. before exposure to the skin wats made. The vapor concentration in both methods was unknown, although in the latter it was practically saturated. For quantitative purposes the following method wits used and in connection with it a special skin applicator was devised.


This consisted of blowing air by means of an ordinary motor blower through concentrated sulphuric acid and calcium chloride, then through a bubbler containing the compound, and connected with a series of glass skin applicators by means of glass tubing as much as possible. The skin applicator consisted of a small glass cylinder of about 1.5 to 2 cm. in diameter an l about 4 cm. long with a small glass handle attached on top, two small tubes at opposite ends for the passage of air laden with the vapors of the compound, and an opening on the bottom for the exposure of skin to the vapor. This opening was 1 cm. in diameter and remained closed until it was established that the vapors were of constant concentration. When the concentration of vapor was constant, the exposure to the skin was made directly for any desired length of time.

The concentration of vapor was determined by dividing the loss of weight of the compound in the bubbler by the total volume of air which was passed through as indicated by the flowmeter. The concentration (nominal) was expressed in terms of milligrams per liter. With certain compounds, the nominal concentrations agreed within 10 to 20 percent by chemical analysis, and this was satisfactory enough for our purposes. The skin irritant efficiency for the vapors of different compounds was judged by comparison of the lowest effective concentrations and the per cent of positive responses to approximately equal concentrations of the vapors, using dichlorethylsulphide as standard.


The compounds were dissolved in suitable solvents, such as absolute alcohol, liquid petrolatum, olive oil, benzene, carbon tetrachloride, and chlorbenzol, for the purpose of determining the lowest effective concentrations, andi for the determination of their skin irritant efficiencies, using two or three solutions which were not effective to four or five which showed positive effects. A definite quantity, usually about 0.02 c.c., was applied to a definite area (1 sq. cm.) of human skin. Larger quantities--that is, about 0.1 c.c. per 5 sq. cm. of skin--were used on doog's skin, so that the concentration per square centimeter of skin was exactly the same for the different species.

However, it was found that so far as the determination of skin irritant efficiency for the purposes of the experiments was concerned the use of the compound in different solvents was not always satisfactory With certain compounds the skin irritant efficiency determined in this way did not agree with those determined by direct application and the different vapor concentrations. This is attributed to differences in volatility, lipoid solubility, coefficient of solubility of compound in skin and solvent, and formation of insoluble and decomposition products. Under the conditions these factors could not be adequately investigated.


Dogs were used principally, owing to the limited amount of human material and the large number of compounds studied. Human skin was used as much as permissible and possible. Monkeys, it horse, and the cock's comb and wattles


were also used. The skin of monkeys responded in the same way as that of dogs. The cock's comb responded principally by local inflammation and necrosis.

So far as known the skin of animals does not blister. Mild effects of irritants on the skin of animals are indicated by simple hyperemia, rashes, moderate swelling and edema, also moderate petechial hemorrhages. Instead of vesication, the more severe effects are characterized by rather extensive edema and swelling, and the severest effects by gangrene and ulceration.

On the other hand, the acute and more severe effects on human skin are characterized by hyperemia and vesication, sometimes pustulation, and the severest effects also by ulceration. The scars in the skins of different species do not differ greatly, and depend upon the degree of ulceration and extent of destruction of tissue. The skin of the Mexican or African hairlesss dog is said to respond to the action of such irritants as cantharides in the same way as human skin, but this was not tried. The part played by the sudoriporous and sebaceous glands in the entry (absorption) of the compounds into the skin was also not studied. In this connection it is interesting to note that horse's skin, which contains sweat glands, was found to be more sensitive to a number of the compounds than the skin of men, dogs, and monkeys. This observation is confirmative of the French.4

Rodents were not used in this work, but the results obtained by others indicate that their response to the various irritants is about the same as that of dog's skin.

No marked differences between shaved and unshaved skin of men were observed, greater susceptibility in animals perhaps being in favor of shaved skin. Great individual variability in both human and animal skins was encountered. It appears that repeated exposure of human skin renders it sensitive and eventually unreliable for further experimentation, although the original lesions may be completely healed. As a rule, pigmentation of skin tended to protect against the irritation by the different compounds. This was frequently tested out in animals by exposing a portion of a large freckle or pigmented spot with neighboring white skin in animals simultaneously to the irritants (vapors and liquid), and almost invariably the pigmented area escaped from the effects entirely. It was suggested by Sollman 5 and observed by Marshall 6 at American University, that the skin of negroes was more resistant than the skin of whites to irritant effects of diechlorethylsulphide.


The following brief summary compiled from a large number of results may be presented at this time:

1. The following compounds were found to be severe irritants as indicated by hyperemia, swelling and edema, ulceration, necrosis, etc., on dog's skin and similar changes, together with vesication on human skin: Arsenic trichloride, bromine trifluoride, chlorisonitrosoacetone, dinitrochlorbenzol (parazol), ethyldichlorarsine, dichlorethylsulphide (mustard gas), iodine pentafluoride, methyldichlorarsine, methyl and propyl "mustard" mixture, methyldibromarsine, phenyldichlorarsine, selenium bromine ethylene derivative, dichlordiethylselenide, "mustard" titanium tetrachloride and di-isothiocyandimethyl ether.


2. The following compounds were found to be mild irritants as indicated by simple hyperemia without vesication, mild urticarial rash; moderate swelling and edema and very little or no necrosis: Normal butyldichlorarsine, O-chlor-chloracetanilide, chloracetophenone, oil from chloracetophenone, chlorethylmethylsulphide, Costa Rica tree sap, b dimethylarsinecyanide, diphenylchlorarsine, diphenylcyanarsine, dichlordiethyltellinochloride, dichlordimethyldithioloxalate, iodoacetophenone, isothiocyanmethylether, isothiocyandimethylether, monochlorethylacetate, monobromethylacetate, selenium chlorine ethylene derivative, selenium ethylene bromine compound, selenium acetylene chlorine compound, dichlordivinylselenide, trichlordiethylselenide, selenium " mustard," trimethylthioarsenite, trimethylarsenite and chlorphenarsazene.
3. The following compounds produced no objective or subjective effects on human and dog's skin: Ammonia silicon tetrafluoride, bromacetanide, benzyl sulfocyanate, ethyl ester of fluorsulphonic acid, "mustard "-mercuric chloride product, juglon, lead tetramethyl, lead tetraphenyl, mercury dimethyl, methyl "mustard, " parabromchloracetophenone, tetrachlordinitroethane and mercury trichlorethylene.
4. As a rule the active arsenicals acted more severely than dichlorethylsulphide during the acute stages. The lesions were more painful, indurated and attached. The ulcers were sharply punched out, clean, and dry, and possessed red bases. Healing occurred promptly. The differences between the different arsenicals were principally quantitative.
5. Dichlorethylsulphide acted more slowly than the aresnicals. The acute effects were less pronounced, and it was more chronic. There was less destruction of tissue in the beginning. Swelling and edema were marked, pouchy and soft in animals. The ulcers were irregular, dirty, purulent and foul. The lesions were generally painless, and secondary infection was common. Healing was slow.
6. As judged by the clinical effects in chronic experiments, lowest effective concentrations of the compounds in solution, and in vapor form, the skin irritant efficiency of the more important severe irritants was about as follows in descending order of efficiency: Dichlorethylsulphide ("mustard gas "), phenyldichlorarsine and methyldichlorarsine. By direct application, dibromarsine and selenium compounds come next in order.
7. The order of protein precipitant power of some of the arsenicals tested agrees in the same direction with their skin irritant efficiency, and it was possible to correlate this in a general way with the quality of skin lesions pro- luced. On the other hand, dichlorethylsulphlide (mustard gas) was the most efficient skin irritant, and its power to precipitate protein was almost negligible, indicating a difference in mechanism of action.
8. Regarding pigmentation of the skin after healing of lesions of several different compounds, the following characteristic features were encountered:
Absence of pigment-dichlordinitrosoacetone.
Faint brown pigment-dichlorethylsulphide.
Deep brown pigment-arsenicals
Metallic gray pigment-organic selenides.
9. These and other differences in the behavior of a number of the compounds that were noted indicate differences in mechanism of action dependent

b Composition unknown.


on differences in chemical structure and composition of the irritants and various physical-chemical and physiological factors. These offer possibilities for further elucidation and correlation of the relations between chemical structure and pharmacological action.

Other possibilities worthy of consideration are the uses of these compounds in the production of experimental lesions and in therapeutics (arsenicals in syphilis, etc.).


(1) Lynch, Vernon: Methods of Testing Skin Irritants. Bureau of Mines Report, No. XV, 48, April 1, 1918. On file, Medical Research Division, Chemical Warfare Service, Edgewood Arsenal, Md.
(2) Lynch, Vernon: 1. Individual Variation in Susceptibility to Mustard Gas (with method of determining the same). Bureau of Mines Report No. XXII, 71, Part 1, July 8, 1918. 2. A Standard Vapor Test for Skin Irritants. Bureau of Mines Report No. XXI, 75, Part 6, July 1, 1918. On file, Medical Research Division, Chemical Warfare Service, Edgewood Arsenal, Md.
(3) Smith, H. W., Clowes, G. H. A., and Marshall, E. K., jr.: On Dichlorethylsulphide (Mustard Gas). IV. The Mechanism of Absorption by the Skin. Journal of Pharmacology and Experimental Therapeutics, Baltimore, Md., 1919, xiii, No. 1, 1.
(4) Mayer, H.: 1. Action of Vesicants on the Skin of Horses, September 10, 1918. (Z 433) 2. Determination of Aggressive Power of Yperite, June 27, 1918. (Z 220) 3. Physiological Action, Preparation and Physiochemical Properties of Offensive Substances, Individuals and Collective Protection, Part 1, Physiology, 11, March 5, 1919. (Z 550) Ministry of Armament and War Manufactures Inspection of Chemical Studies and Experiments. Second Inter-Allied Conference. On file, Medical Research Division, Chemical Warfare Service, Edgewood Arsenal, Md.
(5) Sohlman, Torald: Report of Progress on Prevention, Symtomatology and Treatment of Mustard Gas Burns, Ph. 109, 5, section on "Observations on Symptomatology," July 15, 1918. On file, Medical Research Division, Chemical Warfare Service, Edgewood Arsenal, Md.
(6) Marshall, E. K., jr., Smith, H. W., and Williams, J. W.: Individual Variation in Susceptibility to Mustard Gas (Susceptibility of Negroes). Pharmacological Research Report No. 320, October 31, 1918. On file, Medical Research Division, Chemical Warfare Service, Edgewood Arsenal, Md.