The foregoing experiments would seem to show definitely that the virus of yellow fever, at least when it is present in sufficient amount in the circulating blood of experimentally infected animals, can penetrate the intact skin and produce infection in rhesus monkeys. The successful experiments were made on the abdomen of monkeys where the skin is delicate and apparently highly permeable to the virus. Whether the virus is capable of penetrating the normal and uninjured human skin is not determined by these tests. With regard to the possibility of laboratory infections in man, the hands are of course most likely to come in contact with infectious material. While the skin of the hands is thicker and of denser texture than that of other parts of the body, it is also more subject to injuries and abrasions than other parts. These abrasions, frequently insignificant, often forgotten or even unknown to the person, may therefore serve as portals of entry of this highly active virus in a manner similar to the experiments upon shaven and scarified skin of the monkey. Infection through an injury of this kind possibly took place in the case referred to in the introduction to this paper.
Judging from the results of our experiments with the blood of monkeys nos. 394 and 434 (experiment 2), the danger of infection through the skin would seem to depend largely upon the amount of the virus present in the infectious material.
The virus, at the beginning of the disease, seems to be more concentrated in the blood of experimentally infected animals, as indicated by the following: The amount of blood ingested by a mosquito is undoubtedly very small, and yet in our experience, practically every mosquito that has been fed upon a monkey at the beginning of fever, has become a carrier of the virus. Likewise, in our experiments, infection was produced by smearing only a small drop of blood which in this way came in contact with the skin, must have been exceedingly small, as much of it clung to the hair and was ineffective. On one occasion we titrated the blood of an infected monkey at the beginning of fever, and found that a subcutaneous injection of 0.0001 cc. produced infection and death, the pathology being typical of experimental yellow fever. Smaller amounts have not been tested.
During the course of the disease, the virus decreases apparently in the blood and may even disappear altogether from the blood stream before death. This happening is suggested by a recent case of yellow fever in a European studied. Two lots of normal A. aegypti were allowed to feed upon the patient twelve and thirty-six hours, respectively, after the onset of the illness. Practically all the twelve-hour mosquitoes became carriers of the virus, while the second lot, which had been fed thirty-six hours after the onset of the illness did not become infective. And yet the virus was still present in the blood thirty-six hours after the onset since 5 cc. injected into a monkey induced fatal experimental yellow fever.
One test was made with the blood and organs of a monkey which had died of experimental yellow fever. Other rhesus monkeys were inoculated with heart blood, liver, kidney, spleen and inguinal lymph nodes taken at the necropsy. The monkeys inoculated with liver, spleen and kidney emulsions died, while those inoculated with heart's blood and lymph node remained well. The latter animals (two) when later given 1 cc. of virulent blood from another infected monkey proved immune.
In a recent paper (loc. cit.) it was stated that serum from recovered cases of yellow fever, taken during the period of convalescence possesses power to protect monkeys against experimental virus infection. Hence it seems highly advisable to have a liberal quantity of convalescent serum on hand in the laboratory so that should an accident occur the serum may be administered. One of the writers (J. H. B.) had two accidents in which this measure was applied; in one instance, while taking blood from a native yellow fever patient, his finger was accidentally punctured by the needle used in bleeding, and at another time, he was bitten by a presumably infected mosquito during experimentation. In each case, 5 cc. of convalescent serum was given subcutaneously and no infection occurred. It seems probable that protection was afforded by the convalescent serum.
In performing necropsies on experimentally infected animals, it seems likely that one is handling infectious tissues, and if the animals is killed early in the disease the infectiousness of the material is probably considerable. In the course of over 150 necropsies on experimentally infected monkeys, the other writer (N.P.H.) experienced two accidents. Once a rubber glove broke, and at another time several drops of blood spattered on the forearm and face. In each instance a subcutaneous injection of convalescent serum was given and no ill consequences arose. Because of the rapid deterioration of rubber gloves in the tropics it is the custom to wear two pairs, a precaution justified by finding in several instances a break in one glove and not in the other. Further protection is afforded by the use of rubber sleeves in addition to the usual large rubber apron. Rubber gloves are also worn for bleeding and inoculating animals.