Our knowledge of the proteins of the blood plasma of infants has been very limited, due to the difficulties in methods of study. The small amount of blood available was not sufficient for an exact chemical analysis. The few examinations reported in the literature were limited to the determination of the so-called "colloid lability." It is only since one of us1 published a simple nephelometric method that it has become possible to study the protein fractions in 1 c.c. of blood, and to undertake examinations in a large number of cases. We examined more than one hundred healthy and sick infants. This test is based on the following principle. It has been known since Hofmeister's time that a half saturated ammonium sulphate solution precipitates all the globulins and fibrinogen; a 33 per cent. saturated solution precipitates the euglobulin and fibrinogen; and a 27 per cent, saturated solution will only