COMPOSITION OF THE BLOOD 341 



persons with Pneumonia had a globulin-ratio of from 27 to 50 per cent, 

 and averaging 40 per cent. Other infections showed corresponding 

 increases in the proportion of globulins to total proteins in the serum. 

 Cases of Nephritis gave a high proportion of globulin (24 to 50 per cent.) 

 while those in which nephritis was associated with the accumulation of 

 salts and urea in the blood had also, of course, a high non-protein 

 content. On the other hand a series of patients with diabetes gave 

 normal values for the globulin-ratio excepting in one instance in which 

 a local infection was also present. Individuals afflicted with various 

 types of anemia, hyperthyroidism, goiter, hemophilia, chronic bron- 

 chitis, pellagra, obesity, lead-poisoning, chronic gastro-intestinal dis- 

 orders and neurasthenia presented normal values for the protein-ratio. 

 Exceptionally high values of the proportion of globulin to albumin in 

 the blood-serum, therefore, are associated with Infections or else with 

 Toxemias. 



Globulin expressed 



Non-proteinSj Total proteins, in per cent, of total 



Sample No. Age. per cent. per cent. protein. 



1 .... 27 1.2 7.8 25 



2 . ... 30 1.3 7.4 30 



3 .... 36 1.3 7.3 32 



4 . .".. . . 21 1.3 7.7 26 



5 24 1.1 7.6 16 



6 .... 30 



7 ..... 32 



8 ..-.. 48 



9 . , . . 19 



10 . .". . . 25 



11 . , . . . 48 



12 28 



.2 7.4 32 



.1 8.0 28 



.2 7.9 27 



.2 8.2 27 



.3 7.7 26 



.2 7,3 30 



.3 6.8 29 



13 .... 23 1.2 7.4 24 



14 .... 19 1.25 6.5 29 



15 .... 48 1.25 6.7 31 



16 .... 25 1.3 7.5 21 



17 ..;'-. 26 1.3 6.7 25 



18 .... 29 1.3 6.8 21 



19 .... 26 1.3 7.5 20 



20 ..... 26 1.3 8.2 21 



21 . . . . 26 1.3 8.2 18 



22 26 1.1 7.9 24 



Averages . . . ,'... 1.24 7.5 25.5 



The origin of the rise of globulins in infections is still to be sought. 

 It is not due to, or directly correlated with the development of anti- 

 bodies in the circulation, because as C. L. A. Schmidt has shown, a 

 high degree of immunity to pure proteins may be induced without any 

 rise of globulins. It is not due to the Leukocytosis or increase in white 

 blood corpuscles which often accompanies infection, because Hurwitz 

 and Meyer have shown that the leukocyte-count and the globulin 

 increases do not in any degree run parallel to one another, while C. L. 

 A. Schmidt has shown that the leukocyte-count may be reduced to 

 one-half the normal in rabbits by the administration of Benzole without 

 causing any significant alteration of the globulin-ratio. It is not due to 

 alterations of bodily temperature, because, as Hanson and McQuarrie 



