ALBUMINS OR PROTEINS. 181 



As an example of the formation of a polypeptide by lengthening the 

 chain at the carboxyl end, we may cite the synthesis of leucyl-glycyl- 

 glycine from brom-iso-capronyl-glycine-chloride and glycine-ethyl-ester. 

 The resulting brom-iso-capronyl-glycyl-gly cine-ester is saponified; and the 

 tripeptide, leucyl-glycyl-glycine, results on treating this with ammonia. 

 This, itself, can then be chlorinated, and again united with a peptide- 

 ester, or even with a peptide itself. 



We have gone into the subject of the synthesis of the peptides some- 

 what in detail, owing to the importance of the problem. Synthesis has 

 always been a great factor in biological-chemical knowledge. By this 

 means the constitutions of many substances have been determined, and 

 many debatable questions settled. Synthesis, as we have seen, plays an 

 even more important part in the chemistry of the albumins. With its 

 assistance we hope to determine the constitution of the albumin molecule, 

 and with it, also, we expect to clear up the questions relating to the first 

 decomposition products, the peptones. 



Most of these syntheses have been carried out with inactive amino acids. 

 The structure of these peptides is definitely known, depending on the 

 method of procedure. The subject is not so simple when we consider 

 its stereo-chemical side. We have already mentioned that all the amino 

 acids, excepting glycocoll, contain an asymmetric carbon atom. The num- 

 ber of asymmetric carbons in the polypeptides, therefore, corresponds to 

 the number of amino acids combined in the molecule with the exception 

 of glycocoll. If, for instance, we have a dipeptide of the following general 

 formula, 



NH 2 . CHR . O . CO . NH . CHR . COOH, 



* * 



it is necessary to have, according to van 't Hoff's formula, on account of 

 the two asymmetric carbon atoms, indicated by asterisks, four different 

 active varieties. If we designate the optical antipodes by d and I, the 

 following forms will be possible: dd, II, dl, Id. Two can produce a 

 racemic compound (dd-ll) (dl-ld) . If we start with the racemic amino 

 acids, as has been very generally done, we necessarily expect to obtain 

 two isomeric inactive compounds. This, in fact, is actually the case 

 in practice. Other complications also arise, as, for instance, when we 

 combine a racemic amino acid with an active one; for example, in pre- 

 paring leucyW-tyrosine. Here we have, on the one hand, d/-leucine, 

 and on the other Z-tyrosine. In this case we expect two compounds: 

 a dl- and a ^/-variety. The relations are, of course, much simpler, if we 

 employ only active components in the synthesis. In such a case, we 

 obtain only active peptides ; and if we proceed from those optically 

 active forms of amino acids, which occur in nature, we must obtain amino 

 acid chains, which correspond to those occurring in the albumin molecule. 

 For our requirements the optically active polypeptides are naturally 



