80 CHEMISTRY OF THE PROTEIDS CHAP. 



and as more than 0*2 grm. of this mono-amino-acid is not present in 

 any one albumin, its presence does not vitiate the results of Hausmann's 

 method. Cystin, according to Morner, 1 is not precipitated in a hydro- 

 chloric acid solution by phosphotungstic acid, while according to 

 Winterstein it is precipitated from a dilute sulphuric acid solution. 

 Judging by experiments carried out in Hofmeister's laboratory, 0'05 

 per cent cystin prepared from egg-white is not precipitated by phospho- 

 tungstic acid, if a large excess be avoided, and therefore the cystin-factor 

 may also be neglected as cystin occurs in most albumins in very minute 

 quantities, and even horn-filings, so rich in cystin, do not seem to have 

 given Giimbel any difficulty. 



Sub. 3. Melanins are related to the indol-forming radicals of the 

 albumin-molecule(Samuely 2 ),and therefore to the tryptophane-grouping, 

 and hence Hausmann holds that the melanin which is precipitated by 

 phosphotungstic acid, and which contains 0*16 per cent N. (i.e. a little 

 more than 1 per cent of the total nitrogen) increases by this amount 

 the reading in the di-amino-fraction. It has been shown, however, 

 by Udranszky 3 and Samuely, that nitrogen -.containing humin-sub- 

 stances are formed when carbohydrates are boiled in the presence of 

 amides or ammonium salts, with acids, and therefore, at present, we 

 are not in a position to tell whence the nitrogen in the melanin-fraction 

 is derived, and therefore it is best to follow Osborne and Harris and 

 Giimbel, and to make separate determinations of the melanin-nitrogen. 



3. MONO-AMINO-NITROGEN-DETERMINATIONS 



The values obtained depend on the care with which the di-amino- 

 fraction has been determined. As was pointed out above, some di- 

 amino-nitrogen always remains in solution, and this amount is sufficient 

 to raise the value of the total mono-amino-fraction 1 to 2 per cent, i.e. 

 instead of 60 per cent of mono-amino-nitrogen one may get as high as 

 61 or 62 per cent. 



This mono-amino-fraction is not always readily determined, because 

 Kjeldahl's method is difficult to apply owing to the high percentage of 

 salt present, but it should always be determined, to see whether the 

 total-nitrogen equals the sum obtained by adding up the amid-N + 

 di-amino-N -\- mono-amino-N + melanin-N. 



Giimbel gives for crystalline serum-albumin and edestin of hemp- 

 seeds, casein, keratin, and cartilage the following values, including 

 those obtained by Hausmann and Osborne and Harris : 



1 Morner, ibid. 28. 603 (1899). 



2 Samuely, Hofmeister's Beitrdge, 2. 355 (1902-3). 



3 Udranszky, Zeitschr.f. physiol Chem, 12. 389 (1888). 



