January 3, 1901] 



NATURE 



225 



case of some of the products of digestion, the addition of 

 an acid is found necessary for precipitation. In neutral 

 solution, the concentration of salt necessary for precipita- 

 tion is found to vary with the nature of the proteid. 

 Hofmeister and others have ascertained for a number of 

 proteids the limits of concentration of ammonium sul- 

 phate necessary for their precipitation, and upon this 

 basis have founded the method of fractional precipitation. 

 By this means it has been found possible to separate from 

 a mixture of proteids fairly well defined chemical indi- 

 viduals. 



Another method of separation that is being used to an 

 increasing extent is that of crystallisation ; but, unfor- 

 tunately, its application is somewhat restricted. Dr. 

 Cohnheim has, probably for this reason, relegated the 

 subject to the second part of his book. Egg-albumin, 

 serum-albumin and lactalbumin are the only simple pro- 

 teids of animal origin that have been obtained in a 

 crystalline form. The first method devised is due to 

 Hofmeister. He succeeded in obtaining crystals of egg- 

 albumin from egg-white by first of all precipitating the 

 globulin by half saturation with ammonium sulphate. 

 When the filtrate was allowed to slowly evaporate, egg- 

 albumin gradually separated out in the form of minute 

 globules. By re-dissolving the globules and repeating 

 the process, he ultimately obtained well-defined crystals 

 of egg-albumin which were purified by recrystallisation. 



A much simpler and more satisfactory method has 

 been discovered by Hopkins and Pinkus. After half 

 saturating the egg-white with ammonium sulphate some 

 ammonia is given off, and it was found that, after neutralis- 

 ing the ammonia with dilute acetic acid and then adding 

 sufficient excess of acid to produce a slight precipitate 

 of proteid, the crystallisation of the egg-albumin was 

 rendered much more rapid. The crystallisation induced 

 by this method occurs in closed vessels without any con- 

 centration of the solution, and therefore without the risk 

 of the separation of ammonium sulphate crystals along 

 with the proteid ones. Fifth normal sulphuric acid has 

 been also used, instead of acetic acid, with similar results. 

 Up to the present few attempts have been made to 

 separate different forms of albumin by means of fractional 

 crystallisation ; but, since the work of Hopkins and others 

 has simplified the process, one may hope for farther 

 applications of the method in the future. The method 

 may possibly be found capable of extension to forms of 

 proteid other than the albumins. 



In the succeeding chapter an account is given of the 

 average composition of the simple proteids, and of the 

 methods used in determining their molecular weight. 

 None of the physical methods that have as yet been 

 tested are sufficiently delicate to permit an accurate 

 estimation of the molecular weight of the proteids. 

 Measurements of the lowering of the freezing point and 

 of the osmotic pressure have been tried ; but are very 

 difficult to apply on account of the practical impossibility 

 of obtaining proteid free from admixture with inorganic 

 substances. A full account is given of the chemical 

 methods, which yield more trustworthy results. Although 

 the risks of fallacy are numerous, the results yielded by 

 the chemical methods in many cases render it possible 

 to give at least a minimum value for the molecular 

 weight. 



NO. 1627, VOL. 63] 



The author passes in the next place to a consideration 

 of the chemical characters of proteids. They have the 

 character of potential acids or bases, according to the 

 alkaline or acid reaction of the solution. When one 

 compares the different forms of proteid, one finds that 

 either the basic or acid character may predominate. The 

 simplest forms of proteids, such as the protamines and 

 histones, have a well marked basic reaction. The greater 

 number of the remaining simple proteids play the part of 

 base or acid with almost equal readiness. As a rule, 

 however, the acid character is more emphasised. Simple 

 proteids have, as acids, a distinctly dibasic character. 

 The compound proteids, for example, nucleoproteids, 

 nucleoalbumins and glycoproteids, have a still more 

 marked acid character ; but in their case the acid reac- 

 tion is mainly due to another organic group that has 

 united with a molecule of a simple proteid. 



In addition to these salt-like compounds of albumin, 

 one finds compounds with inorganic material, for example, 

 the halogens or iron, in which the halogen or iron is 

 present in a more stable organic combination, and not as 

 an ion. Some of these have been prepared artificially ; 

 others by means of vital processes. It appears certain 

 that the organism is capable of forming a stable organic 

 compound of proteid containing iodine in which the char- 

 acter of iodine as an ion is lost, even when the only sub- 

 stances in the food given contained iodine in an inorganic 

 form. 



The chief characteristic reactions of proteids next con- 

 sidered may be divided into precipitation and colour re- 

 actions. In virtue of their basic character, most proteids, 

 when in the presence of an acid, may be precipitated by 

 the precipitants of the alkaloids. As acids, proteids form 

 insoluble salts with most of the heavy metals. On these 

 facts, methods have been founded for the estimation of 

 the basic or acid equivalents of various proteids. Cohn- 

 heim illustrates this by the following example. The 

 hydrochloric acid salt of a proteid and calcium phospho- 

 molybdate = the insoluble compound of the proteid with 

 phosphomolybdic acid and calcium chloride. Since the 

 hydrochloric acid salt of a proteid reacts as an acid to 

 phenolphthalein while calcium chloride is neutral, the 

 diminution of acidity after precipitation can serve as a 

 measure of the basic equivalent of the proteid. The 

 more basic forms of proteid, for example, histones and 

 protamines, are precipitable by reagents for the alkaloids 

 from neutral or even alkaline solutions. 



The colour reactions of the proteids are dependent on 

 the presence of certain organic radicles in the proteid 

 molecule, and owe their importance to the light which 

 they throw upon the chemical structure of the proteids. 

 The failure or success of any given colour reaction indi- 

 cates the absence or presence of the corresponding 

 organic group in the proteid molecule. By a careful 

 study of the colour reactions and of the simpler decom- 

 position products of the proteids, it may be possible in 

 the future to subdivide the different proteids into struc- 

 turally distinct classes, and thus to establish a classifica- 

 tion of the proteids upon a chemical basis. 



The most important of the colour reactions is the biuret 

 reaction. The conditions necessary for its occurrence 

 have been very fully studied by Schiff. It is given by all 

 chemical bodies containing two CONHj groups united 



