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

Communicable Diseases, Table of Contents




Infectious jaundice is not a new disease, epidemics having been reported upon as far back as the first half of the eighteenth century.1 In 1914 certain Japanese investigators isolated a spirochete from cases in their own country to which they gave the name Spirochæta icterohæmorrhagiæ.1 The intermediate host of the parasite is the brown rat. This spirochete is to-day generally accepted as the specific cause of the disease.

The United States Army reported a total of 452 cases of spirochetal jaundice during the period of the World War, with a resultant loss of 9,251 days and 15 deaths. The distribution of these cases by countries is indicated in the following table: United States, 279; Europe, 108; Philippine Islands, 15; Panama, 9; other countries, 5; transports, 1; total officers, 35; total Army, 452. To these 452 original admissions must be added 80 instances in which the same malady occurred concurrently with other diseases, making a grand total of 532 cases.


Typhus fever has been long known, and the World War added little, if anything, to clinical knowledge concerning it. Its transmission by the louse is accepted. Its incidence in the Army during the World War was as follows: United States, 15; Europe, 7; other countries, 19; officers, 1; total, 42. To the 42 original admissions for typhus fever must be added 5 more cases in which this malady appeared as a concurrent disease. Three of these individuals died and the disease was responsible for 1,335 days lost.

Although the war did not, properly speaking, advance in measurable degree the clinical, epidemiological, and etiological phases of the available knowledge concerning typhus fever, nevertheless, the startling epidemics of the disease which occurred in Russia, Poland, and other countries during the war period served to stimulate the labors of the Typhus Research Commission of the League of Red Cross Societies to Poland, the results of whose investigations became available in the main report of this commission which appeared in 1922.2 This commission looked upon "the determination of the exact nature of the specific cause of the disease * * * as the most important goal."2

The transmission of typhus fever by the louse was accepted at the start. The one organism found to be most definitely and uniformly demonstrable in lice that had fed upon typhus patients was Rickettsia prowazeki. This Rickettsia was identified in each instance upon its appearance as observed in serial sections of lice. It was found in the lice of 27 out of 52 experiments.3 Changes in technique based upon experience secured positive results in practically every one of the last third of the experiments.

aUnless otherwise stated, all figures for the World War period are derived from sick and wounded reports sent to the Surgeon General.-Ed.


Rickettsia pediculi was found occasionally. Microorganisms other than Rickettsia did not appear in any of the 52 experiments in which lice were nourished upon typhus patients. Animal experimentation designed to prove Rickettsia prowazeki the specific cause of typhus fever is reported in detail. The authors' conclusions are as follows:4

The presence of Rickettsia prowazeki in lice in our experience is proof of the presence of the virus of typhus.

A variable percentage only of lice nurtured upon typhus patients acquire the virus of typhus; and this holds true in boxes where all lice have equal opportunities to become infected.

After allowing for the technical difficulties in making adequate search for Rickettsia and injections from the same louse and for the uncertainty of the reaction of guinea pigs to typhus blood in the test for immunity, we believe that the data from the above experiments are sufficient proof that the virus of typhus and Rickettsia prowazeki are inseparable.

Concerning Rickettsia, it is stated5 that this "is the group name given by da Rocha-Lima (1916) to minute microorganisms with certain peculiarities found in lice. The name honors the memory of Howard Taylor Ricketts, who first described microoganisms possibly of this type in connection with studies upon typhus (Ricketts and Wilder, 1910)." The present knowledge of Rickettsia is summarized as follows:6

A satisfactory definition of Rickettsia is not possible at present. The properties in common of the 13 or 14 microorganisms so far described under this name are as follows:

Morphology.-Bacteriumlike on the whole. They are smaller than bacteria and occur characteristically in pairs. Large forms, bacillary and filamentous, have been described in connection with two carefully studied Rickettsias-Rickettsiaprowazeki and Rickettsia lectularius-and it seems probable that a simple cycle or sequence in morphological development is a characteristic of the pathogenic forms.

Staining reactions -Difficulty of staining with the common staining solutions used for bacteria is a striking feature, as is the failure to retain the stain by Gram's method. The only satisfactory staining methods are the modifications of Romanowsky's method; of these, the most satisfactory is Giemsa's solution.

Mobility -Motileforms have not been seen.

Cultivation -So far all have resisted cultivation with the exception of the Rickettsia from the sheep louse. It grows on a relatively simple glucose blood agar medium.

Resistance to physical and chemical agents -Not enough work has been done to generalize. The viruses of typhus (da Rocha-Lima, 1919, p. 240) and Rocky Mountain spotted fever (Wolbach, 1919) are extremely susceptible to heat, drying, and chemical agents. On the other hand, the virus of trench fever resists 80° C. of dry heat for 20 minutes and drying for many months (Byam and Lloyd, 1919).

Host specificity -All Rickettsias have insect hosts which in the case of the pathogenic ones are the vectors. All are highly specific for their insect host, while the pathogenic ones may infect widely separated mammals.

Hereditary transmission -In every instance where careful study has been made it has been found-with the exception of the Rickettsia of typhus-that the organisms pass down through successive generations in the eggs. Da Rocha-Lima has offered some evidence that this is also true of Rickettsia prowazeki, and Sergent, Foley, and Vialette, 1914 (quoted by Nuttall, Parasitology, vol. 10), accidentally communicated typhus to a monkey and a man with the offspring of lice which were supposed to be infected only with relapsing fever.

Classification is of course impossible, and it is probable that we have already included under Rickettsia a number of very different microorganisms. The Rickettsia of the sheep louse has little to distinguish it from bacterium; yet we believe the Rickettsia of typhus has a number of peculiarities which necessitate its separation at present. The Rickettsialike


cause of Rocky Mountain spotted fever, which we prefer for the present to consider under a distinctive name, while resembling in many ways Rickettsia prowazeki, is very unlike the morphologically simple Rickettsia of trench fever.

At the present the opinion seems generally held that Rickettsia prowazeki is the specific cause of typhus fever. This knowledge may be credited to the epidemiological opportunities indirectly afforded by the World War.


In 1915 there made its appearance among British troops on the Western Front a disease7 which came to be known as "Trench fever"8 and which gradually was recognized as a specific infection. The armies in Salonika likewise reported cases of this disease.9 There is no available evidence indicating that this disease had ever been recognized as a clinical entity before 1915.


The following figures, which represent the occurrence of trench fever in the Army, include a number of experimental cases, but do not embrace relapses from either experimental or natural causes. The total number of primary admissions was 798, divided as follows: Officers, 54; white enlisted men, 531; colored enlisted men, 2; color not stated, 211. All of these cases, except 12, occurred in our forces in Europe; the 12 cases were admitted in the United States. To the 798 primary admissions are to be added 103 cases which were concurrent with other diseases for which admission was made, thus giving a total of 901. Among these, there were 2 deaths, 1 white enlisted man and 1 color not stated. Both deaths were in Europe.

The foregoing figures suggest a point of interest as regards the racial distribution of this disease. The admission rate per 1,000 was, for white enlisted men in the whole Army, 0.15, as against 0.01 for colored enlisted men; and in Europe 0.35 for white enlisted men and 0.02 for colored enlisted men. Assuming the probability that essentially the same conditions as to louse infestation obtained in both white and colored combat troops, this notable difference in race incidence would seem at first glance to indicate a relatively higher immunity on the part of the colored man. On the other hand, colored troops had relatively much less service in the trenches than did white troops, and it is probably much nearer the truth to assume that lice infected with the virus of trench fever were largely, if not wholly, confined to the combat areas.


The researches of the trench fever research committee of the American Red Crossb whose report was issued in 1918 afforded at the time of publication the last word on the etiology and transmission of this disease.10 For the investigation of the problems concerning the etiology and transmission of the disease, human subjects were necessary, since the disease was not transmissible to animals. The consent of the commander in chief, A. E. F., to the use of soldiers

bThe members of the commission were as follows: Maj. Richard P. Strong, M. C.; Maj. Homer F. Swift, M. C.; Maj. Eugene L. Opie, M. C.; Capt. Ward J. MacNeal, M. C.; Capt. Walter Baetjer, M. C.; Capt. A. M. Pappenheimer, M. C.; Capt. A. D. Peacock, R. A. M. C. (T); and Lieut. David Rapport, M. C.


who might volunteer for this human experimentation was secured by the chief surgeon, A. E. F. Out of hundreds who volunteered, 82 were selected. All these volunteers were subjected to detailed physical examination to exclude any unfit, and bacteriological examinations were then made of the blood, urine, and feces to eliminate those who might be suffering from chronic infections, and carriers. The whole detachment was strictly segregated, and the most complete records as to temperature and condition of skin and clothing were made. A semiweekly bath and weekly sterilization of clothing were a part of the routine.

In seeking the specific etiological factor, the first step was an inquiry into the possibility of infection with any members of the enteric group of bacteria. Briefly, the examinations of blood, urine, and feces in cases of spontaneous and experimental trench fever by methods adopted in the search for the typhoid and paratyphoid bacilli were consistently negative. Serological reactions gave no indication that any of these organisms was culpable. Structures resembling spirochætes had been previously found by at least one worker on the hematology of trench fever.11 He looked upon the possibility of their being the causative factor in the disease as not incompatible with the filterability of its virus since "some spirochætes are known to be filterable * * *." Otherwise the search for spirochætes in the blood (as conducted by the members of the Red Cross commission by the method of anaerobic cultures) was entirely unsuccessful. The Wassermann test likewise was consistently negative, and thus failed to give any suggestion that trench fever might be a modality of syphilis, or due to any spirochæte of close biological relationship to Spirochæta pallida. The commission confirmed the earlier experiments of McNee, Brent, and Renshaw that the disease was infectious and transmissible by the blood. Thirty-four volunteers who previously had been studied with great care were inoculated with blood or some constituent portion thereof taken from trench fever patients during the febrile paroxysms. Of these, 23 contracted the disease with an incubation period of from 5 to 20 days. Careful consideration of the results of experimental inoculation brought the commission to the conclusion that "The virus or organism of trench fever is present particularly in the fluid portion of the blood, and is not contained within the blood corpuscles themselves."

Investigation of the filterability of the virus resulted in proof that "at least one stage of the development of the virus of trench fever is filterable and ultra-microscopic," though elsewhere it is stated that the "virus is not filterable with ease in centrifuged plasma or serum."

Concerning the thermal death point of the virus, it was found that the virus resists a temperature of 60° C. moist heat for 30 minutes, and is fully virulent after such treatment, but is killed by a temperature of 70° C. moist heat for 30 minutes. Obviously, therefore, a temperature of 55° C. for 30 minutes, which destroys the Pediculus humanus and its ova, does not suffice to destroy the virus of trench fever which may be present on the underclothing of trench fever patients. Furthermore, its stability is perhaps one of its most striking characteristics. Immediate suspension of the thoroughly dried virus in a large volume of saline solution for several hours does not attenuate its virulence, and it has already been pointed out that it resists drying in the urinary


sediment of trench fever cases and in louse excrement. For these reasons the organism causing trench fever may apparently be most appropriately classified as a resistant filterable virus. The virus is invariably present in the blood in the early stages; it is not present in the feces; it is present in the urine; sometimes it appears to be present in the mixture of saliva and sputum, as indicated by successful inoculation experiments.

Thirty-eight of the 82 volunteers were employed in experimental investigation of the transmission of trench fever by the louse. A pure-bred strain of lice was obtained from the Lister Institute for this purpose. The so-called "box method" of handling these lice was made use of. When lice were to be allowed to feed upon trench fever patients the small, round, cardboard box in which they were confined was placed, after removal of the cover, upon the surface of the skin of the forearm. The period of feeding was in each instance not less than 30 minutes and was repeated three times a day. Lice were placed upon healthy volunteers at varying intervals of time after removal from the patients, with intent to exclude direct mechanical transmission through the medium of the parasite's biting mouth parts, and as well to obtain information with regard to the length of time the louse might remain infective. These experiments led to the conclusion that the disease "is transmitted naturally by the louse, Pediculus humanus, Linn., var. corporis, and that this is the important and common means of transmission; that the louse may transmit the disease by its bite alone (the usual manner of infection), or the disease may be produced artificially by scarifying the skin and rubbing in a small amount of the infected louse excrement." Also "that a man may be entirely free from lice at the time he develops trench fever, the louse that infected him having left him some time previously as its host, and that the louse need only remain upon the individual for a short period of time in order to infect him." It was furthermore shown that no evidence could be obtained pointing to the hereditary transmission of the virus of trench fever in the louse; and finally there is evidence that, if the virus undergoes development in the louse, it requires 6 to 10 days to do so; there is a little evidence that suggests the minimum incubation period to be about 4 days and that "lice may remain infected for at least 10 days and possibly 13."

The most suggestive discovery in all attempts to identify the specific cause of this disease has been Rickettsia bodies. Swift12 states that "while it is difficult not to believe that there is a causal relationship between the virus of trench fever and the Rickettsia bodies, it will be difficult to establish definitely such a relationship until it is possible to obtain pure cultures of the bodies and with them to reproduce the disease. In this connection it must be recalled that the relation of Rickettsia bodies to other microorganisms has not been established. They may be specific microorganisms; they may be a granular stage through which some other microorganism is passing; or, finally, they may be cell inclusions, the result of the action of some invisible virus on the cell protoplasm, and thus resemble the Guarnieri bodies in vaccinia, the Negri bodies of rabies, the molluscum bodies in molluscum contagiosum, and the cell inclusions in trachoma." Ledingham13 succeeded in producing agglutination of Rickettsia in emulsions prepared from dried lice excreta by the use of immune


sera from four experimental rabbits and one guinea pig. He found, however, that agglutination disappeared beyond a dilution of 1 in 40.

From critical consideration of the foregoing it would seem that the final conclusion as to the specific cause of trench fever must be that it is not yet indubitably known, although available evidence now points more suggestively to a Rickettsia than in any other direction.


The period of incubation in louse-borne trench fever varies from 14 to 30 days, though the suggestion is offered that this might be shortened in cases of infection resulting from large amounts of virus.10 Certain vague prodromata-headache, fever of a low grade, and pain in the extremities-are complained of by a minority of individuals. Otherwise the onset is sudden and characterized by dizziness, headache, retrobulbar pain, particularly on movement of the eyeballs, nystagmus when the eyes are directed to either side, conjunctivitis, and a sudden elevation of temperature to 103° or 104° F. The febrile reaction varies much in its characteristics. It may last about a week, to be followed by a period of defervescence and a short relapse, or it may persist for several (often six) weeks and be marked during that period by indefinite relapses; and, finally, it may assume quite distinctively the form of a regularly relapsing fever with apyretic intervals lasting six or seven days. A less frequent picture is that of a low continued fever which persists, with only slight remissions, or none at all, from one to two months. The skin in three cases out of four presents an eruption consisting of erythematous spots or papules, most intense on the ventral and dorsal surfaces of the torso. The individual lesions average 3 or 4 mm. in diameter, are pink in color, and the color disappears under pressure. The period of their duration is short, often no more than 24 hours. Most characteristically they first appear during the initial stage, but they are sometimes first seen during a relapse.

The blood picture is variable. Many cases show a moderate leucocytosis (13,000 to 17,000), which recurs with each relapse. On the other hand, in certain cases the leucocytes are normal, while a few manifest a leucopenia (occasionally as low as 3,500). The urine frequently contains albumin in small amounts, but there is no other evidence of true inflammation of the kidney. The spleen is enlarged in a considerable majority of instances.

The most impressive subjective symptoms of trench fever are pain and tenderness. These are referred particularly to three regions or systems-the bones, the head, and the muscles. The "shin pains" are present in 75 per cent or more of all patients. They are boring or lancinating in character and increase in severity so much at night as seriously to interfere with sleep. They appear most characteristically on the third day of the disease or later, but in certain cases do not occur until the first or second relapse. They are accompanied by marked tenderness of the tibia to pressure. Pain of a similar character occurs in the scapula in many cases. Joint pains occur with a frequency equal to that of shin pains in both the upper and lower extremity. The headache is a universal symptom. It is commonly frontal or postorbital, though it may be generalized, and it lasts for two or three days, usually recurring with


each relapse. It is accompanied by a peculiar tenderness to pressure in the supraorbital region which occasionally involves the entire scalp.

Muscle pain is complained of in the lower extermities, in the abdominal wall (either localized or generalized), in the lumbar region where it occurs in 80 per cent of all cases, in the muscles of the shoulder girdle, and in the cervical muscles. It is accompanied by tenderness on palpation.

Anorexia and coated tongue are the most prominent of the gastrointestinal symptoms.

Although the pain is so constant and so marked and the reflexes are exaggerated, it is not believed that the central nervous system is directly attacked by the virus of trench fever. The nervous manifestations are probably no more than may be accounted for on the basis of toxemia.

The pulse usually parallels the temperature in the first stages of the disease, but later shows a tendency to acceleration. The dyspnea, tachycardia, precordial pain, increase in the size of the heart, all indicate marked involvement of that organ. It has been assumed either "that trench fever virus has a selective action on the heart muscle such as we see in rheumatic fever or in the specific infiltration in syphilis 'of the heart,' or that the 'toxin in trench fever acts upon the heart muscle in a similar way to that seen in pneumonia, typhoid fever, influenza, bronchitis, or other acute infections.'"10


The most important of these is concerned with the heart, and has been variously termed "effort syndrome," "disordered action of the heart," and "tachycardia." The circulatory manifestations of trench fever have already been described. The persistence of the indicated cardiac condition after the apparent cessation of activity is probably to be ascribed-at least, in large measure-to the desire of medical officers to return soldiers to duty as soon as possible. Convalescence appears to be essentially a lengthy procedure in this disease, as in dengue fever, and acceptance of such a view in the management of convalescence will allow for complete recovery without the appearance in any marked degree of this cardiac disorder. Thus, the American Red Cross commission, previously quoted, states that-

Among our patients, we feel that up to the present time none have shown a condition of D. A. H. after the infection was overcome. This probably is due to the fact that our subjects were carefully chosen, and those who had previously shown symptoms of cardiac weakness were not inoculated. The subjects were all young and strong, and at the time of inoculaton were not suffering from other infections, nor had their resistance been lowered by long duty in the trenches or exposure to other forms of strenuous work * * * On the other hand, the absence of permanent effect upon the heart may have been due to the opportunity we had of holding the patients until we felt they were fit for active duty.

Except for the cardiac complications and sequelæ of trench fever, the concurrent diseases in the Army were widely various, and were not such as to indicate a pathological relationship between themselves and the original infection.



Since the disease uncomplicated is never fatal, and since it is not transmissible to animals, the pathology is a sealed book. The clinical evidence of changes in the heart and spleen, as well as the blood findings, have been referred to under symptoms.


Diagnosis is to be made upon the symptoms and signs hereinabove described, of which the shin pains and shin tenderness are perhaps the most essentially characteristic.

From influenza, trench fever may be differentiated by the absence of respiratory symptoms and signs, by the characteristic pain and tenderness, by the relapses, by the splenic enlargement, and by the eruption.

Typhoid and paratyphoid fevers are of gradual onset, are accompanied by certain digestive disorders, present a spleen which is less enlarged (if at all) and softer than is the case in trench fever, lack the characteristic pains and tenderness, show a leucopenia, and give a blood serum capable of agglutinating the causative organism in high dilution. Bacillus typhosus and Bacillus paratyphosus may also be recovered from the blood, feces, and urine.

Trench fever and malaria differ very characteristically in their temperature charts, in the matter of febrile paroxysms, in the skin eruption, pains, and tenderness which are present in the former disease, and in the absence of the malarial parasite in the blood of trench fever cases unless the two diseases coexist.

Relapsing fever, because of its mode of onset, its pain, and rash, may be confused with trench fever; but a crisis on the seventh day with a relapse at the end of another seven-day period in relapsing fever indicates a difference between the febrile processes. Both liver and spleen are enlarged in relapsing fever, the spleen alone in trench fever. The causative spirochetes are present in the blood of relapsing fever and absent therefrom in trench fever. Salvarsan exerts a marked effect upon relapsing fever, but none on trench fever.

Dengue fever may suggest trench fever in its mode of onset and distribution of painful areas, but the acute stage of dengue is short, with an intermission occurring on the third to fifth day. The rash in dengue fever is erythematous or scarlatiniform during the first paroxysm and measleslike in the second paroxysm. Dengue is invariably characterized by a marked leucopenia; trench fever presents a moderate leucocytosis in a great majority of cases.

In typhus fever the onset is more gradual than in trench fever and is accomplished in successive steps. Toxemia becomes increasingly more profound as typhus fever progresses, while the toxic manifestations of trench fever-such mild ones as there may be-are more intense within the first few days after onset and rapidly subside. The characteristic relapses of trench fever are not found in typhus fever. The skin eruption in typhus appears on the third to the fifth day and is macular in character, changing to petechial. Typhus fever has a high mortality; trench fever is never fatal.

Malta fever is to be differentiated from trench fever, first, by the different temperature curve, by the absence of the eruption and characteristics pains


and tenderness, by recovery of the causative organism from the blood and urine, and by serological methods.

There is, however, no serological or other laboratory procedure which is specific for the diagnosis of trench fever.


General preventive measures during the war consisted essentially in the delousing of officers and men, together with their effects, a detailed description of which is in Volume VI, Sanitation.

No attempts, apparently, were made to attenuate by heat the virus of this disease as it occurs in louse excrement and in the urine of trench-fever patients, and to vaccinate experimentally with such material. Since the infection of laboratory animals is not possible, such experimental attempts at vaccination could hardly lead to results of practical value for the reason that the source of such vaccine could only be trench-fever patients themselves, of whom a very considerable number would undoubtedly be necessary to supply material sufficient in amount for large bodies of troops. Again, no reference can be found to the attempted protection of noninfected individuals by the use of serum from convalescent patients.

The relatively small number of cases of trench fever reported for the whole United States Army can not bespeak efficiency of the preventive measures in operation among American troops, in view of the fact that the major portion of our combat troops were louse infested at the time of the signing of the armistice. But other considerations must be taken into account. A study of the statistics of trench fever in the Third Army shows that three-fourths of all its cases occurred in two divisions, as follows:14

Incidence of trench fever in troops of Third Army, January 1 to March 1, 1919





42d Division 


Third army troops


2d Division


Third Corps troops


1st Division


Fourth Corps troops


52d Division





4th Division


3d Division


Again, the Third Army alone reported within only two months (January 1 to March 1, 1919) nearly a third of all the cases in the whole American Expeditionary Forces for the entire period of the war. That these figures represent the actual state of affairs is difficult to believe. They tend to suggest rather that in reality the cases in the Army exceeded the reported number of 901, and that many diagnoses were missed either because of transfer of the infected individuals from one station to another or-more especially-"because medical officers were not acquainted with the manifestations of the disease."14 This impression is definitely reinforced by a consideration of the far greater uniformity with which trench fever is shown to have invaded the various units of the British Expeditionary Force.



No specific method of treatment is available. Salvarsan and the other antisyphilitic arsenicals are without effect. A symptomatic therapy consisting of complete rest in bed during the infectious period, a diet of good nutritive value, and the exhibition of such drugs as aspirin and Dover's powder in doses sufficient to control the pains are indicated. The most important factor in treatment is the recognition of the necessity for prolonging the convalescent period until the cardiac condition and action have become entirely normal. Patients should be kept in bed until all probability of a relapse is at an end. The amount of time allowed out of bed should then begin with a few hours and gradually be increased daily, provided no return of the cardiac symptoms is noticed. As soon as the patient reaches the point of remaining up and about all day, guarded and carefully supervised exercises are to be begun. These must be carefully controlled and the appearance of dyspnea, cardiac palpitation, dizziness, pain, fatigue, headache, excessive increase in pulse rate, and cyanosis are the signal for moderation in the amount of exercise taken. The exercises employed are of two kinds: Setting-up exercises and practice marches. As soon as the patient can endure a practice march of 5 miles with full field equipment and return therefrom without evidence of undue weakness his convalescence is looked upon as completed


(1) Jobling, James W., and Eggstein, A. A.: The Wild Rats of the Southern States as Carriers of the Spirochæta Icterohemorrhagiæ. The Journal of the American Medical Association, Chicago, 1917, lxix, 1787.

(2) Wolbach, S. Burt; Todd, J. L.; and Palfrey, F. W.: The Etiology and Pathology of Typhus. (Report of the Typhus Research Commission of the League of Red Cross Societies to Poland). Harvard University Press, Cambridge, Mass., 1922, 3.

(3) Ibid., 43.

(4) Ibid., 112.

(5) Ibid., 116.

(6) Ibid., 123-4.

(7) Graham, J. H. P.: On a Relapsing Febrile Illness of Unknown Origin. The Lancet, London, September 25, 1915, ii, 703.

(8) Hunt, G. H., and Rankin, A. C.: Intermittent Fever of Obscure Origin. The Lancet, London, 1915, ii, 1133. Also, Hunt, G. H., and McNee, J. W.: Further Observations on Trench Fever. Quarterly Journal of Medicine, Oxford, 1915-16, ix, 442.

(9) Hurst, A. F.: Trench Fever. Journal of the Royal Army Medical Corps, London, 1917, xxxviii, 207.

(10) Report of Commission of American Red Cross Research Committee on Trench Fever. Oxford University Press, 1918.

(11) Coles, Alfred C.: Spirochætes in the Blood in Trench Fever. The Lancet, London, March 8, 1919, i, 375.

(12) Swift, Homer F.: Trench Fever. Archives of Internal Medicine. Chicago, 1920, xxvi, 76.

(13) Ledingshaw, J. C. G.: Agglutination Experiments with Trench Fever Riekettsia. The Lancet, London, June 12, 1920, i, 1264.

(14) Swift, Homer L.: Trench Fever in the American Expeditionary Forces. The Journal of the American Medical Association, Chicago, 1919, lxxiii, No. 11, 807.