﻿ETIOLOGY OF DENGUE FEVER. 97 



of communicability is greatest and that on or about the tenth day, when the 

 eruption has disappeared and after earbolie-acid baths, the danger of communica- 

 tion ceases." We think that this observation is probably correct, but explainable 

 quite apart from any belief in direct contagion. 



Fayrer quotes many Indian medical officers who express most diverse views 

 concerning the epidemiology of the disease, yet nearly all seem to be readily 

 accounted for under a theory of mosquito transmission. 



Manson (16) does not express himself as believing in any special method of 

 propagation, though he quotes Graham's work. 



Van der Burg (17) says that epidemics are most common in the hot, moist 

 season, and cease when the weather grows cooler, especially if very heavy rains 

 occur, accompanied by cool winds. He thinks that dengue is very probably 

 contagious, but that the contagious principle is different in character from that 

 of smallpox, epidemic influenza and like diseases. He states that unknown 

 exogenous conditions are necessary for its spread, wherefore many regard it as 

 a "contagious-miasmatic disease." 



Scheube (18) says that the question of contagion is still undecided, but he 

 considers the disease one requiring unusual conditions for its spread. 



Agramonte (19) expresses the belief that the disease may exist in a latent 

 state for weeks, under conditions not as yet well recognized. 



Stitt (20) says that the disease is not infectious in the same sense as is 

 influenza. 



Harris Graham (21) in 1903 expressed his definite belief that the disease is 

 transmitted by mosquitoes, and cited some interesting experiments in support of 

 his belief. This belief seemed well founded, and the experiments in support of it 

 almost conclusive, but the value of his work was impaired by the fact that in 

 the same paper he described as the etiologic factor an intracorjrascular "organism 

 with amoeboid movement, and in many ways resembling the Plasmodium malariw." 



It is probable that many workers, having satisfied themselves that such an 

 organism did not exist in the blood of dengue, concluded that Graham had worked 

 with some other disease, and so did not recognize the real importance of his work. 

 We are unable to believe that the disease is characterized by the presence of any 

 organism microscopically resembling that of malaria, but we think that our 

 observations do lend support to Graham's conclusions as to mosquito transmission. 



Carpenter and Sutton ( 22 ) failed to transmit the disease by means of mosquito 

 bites, but we think that we can explain their failure more readily than we can 

 the negative results which we obtained in some of our own experiments. 



2. THE FORT WILLIAM M'KINLEY EPIDEMIC. 



The epidemic from which we drew our cases is of great interest, as 

 we think that its history fully supports our belief that the disease is 

 mosquito-borne. Port William McKinley is situated about 5 miles from 

 the city of Manila, upon rolling land consisting of slight elevations 

 interspersed with low, damp country. There are stationed at this post 

 two regiments of infantry, two squadrons of cavalry, one battery of field 

 artillery and one company of engineers. 



The infantry barracks are situated near a small stream which drains 

 the lowest part of the post in their vicinity. This stream is overgrown 

 with a rank tropical vegetation and is an ideal breeding-place for 

 mosquitoes. The epidemic began in the barracks of the Sixteenth In- 

 fantry, which is situated nearer this stream than any of the other barracks 



