It should be pointed out that following inoculation of liver emulsion, and sometimes even of blood, it is impossible to tell whether an early fever resulting is due to a protein reaction, to bacterial infection, or to true yellow fever. If plenty of animals are available the only safe plan is to take blood for subinoculation and for cultures, and to feed mosquitoes. Blood may be kept frozen for some time before inoculation if it is found to be necessary, or if it is desired to await the outcome of the disease in the animal bled.

A combined blood and liver transfer showed the presence of virus in rhesus No. 46 twenty days after the original inoculation with kidney from the autopsy in the case of J. V. 0. The same animal had had an infected body wall and a bacteriemia.

On the basis of later experience we can see some of our mistakes. In the second S. R. passage (rhesus No. 133) blood transfer should have been made at the time of initial fever, the day following inoculation. In the case of rhesus No. 60, mosquito infection from J. V. 0., transfer probably should have been made when the temperature reached 103.9°F. on the fifth day. Two days later, with a fever of 104.4°F., the blood appeared to be non-infective. A temperature of 104°F. in monkeys is usually a safe borderline between fever and no fever, but seemingly not always.

Our experience with the South American viruses so far indicates that it takes much care, patience, and an abundance of monkeys to build up and maintain a high degree of virulence.