ALBUMINS OR PROTEINS. 161 



that the mucins and their related substances can contain as much as 30 per 

 cent glucosamine. It is not at all unreasonable to expect varying results 

 from the same mucin, because it is absolutely impossible to purify these 

 compounds thoroughly. It is more striking that egg-albumin, which 

 is so easily crystallized, does not give concordant results on the amount 

 of glucosamine. It must also be remembered that egg-albumin contains, 

 besides albumin and globulin, ovi-mucoid, which contains about 30 

 per cent glucosamine. By the recent investigations of Fr. N. Schulz 

 and Zsigmondy, 1 it was shown how extremely difficult it is to free 

 egg-albumin from colloidal substances, even after sixfold crystalliz- 

 ation. In recrystallized egg-albumin, values varying from 16 to less than 

 one per cent of glucosamine were found. 2 We do not err, in assigning the 

 fluctuations of the carbohydrate content, which far exceed the analytical 

 error, to this cause of varying purity of the substance under investiga- 

 tion. It is, therefore, not yet determined whether egg-albumin is entitled 

 to a " carbohydrate group." The same holds true regarding serum 

 albumin. This, also, invariably incloses some serum-mucoid, 3 which is 

 relatively rich in glucosamine. In fact, serum-globulin is distinctly 

 different, in that, aside from a trace of glucosamine, it also splits off 

 grape-sugar. Now serum-globulin is obtained by precipitation with 

 ammonium sulphate solution. To purify this, to the extent possible 

 with the crystallizable proteins, is entirely impossible. Besides grape- 

 sugar, serum also contains small quantities of other carbohydrates of 

 unknown constitution. It is very probable that the precipitated serum- 

 globulin contains such a carbohydrate mixed with it. Up to the present 

 time, we have had no investigation which would warrant us in assigning 

 any nitrogen-free carbohydrates to the albumin molecule. 



If we sum up what we know about the " carbohydrate groups " 

 of the proteins, we will conclude that the mucins and mucoids contain 

 such groups; while although the remaining proteins may contain carbo- 

 hydrates, their presence has not been proved positively. 



It is very important that absolute clearness should prevail in regard to 

 this question. We shall see later, that many facts make it seem probable 

 that carbohydrates are formed from albumin. The assumption that, 

 according to our present knowledge, the " carbohydrate group " of the 



1 Fr. N. Schulz and Zsigmondy: loc. cit. 



2 E. Abderhalden, P. Bergell, and T. Dorpinghaus: Z. physiol. Chem. 41, 530 (1904). 

 Direct experiment showed that repeated recrystallization reduced the quantity of 

 glucosamine present in albumin. Albumin crystallized once gave 7 per cent, three 

 times gave 4 per cent, while the seventh time showed only 2.5 per cent glucosamine. 

 That the values in this case are higher than in the work just mentioned, is due to the 

 fact that the crude osazone was weighed, while in the former the analytically pure 

 osazone was used as a basis for the calculation. 



3 C. N. Zanetti: Ann. Chim. Farmac. 12, 1897. 



