26 CHEMICAL CONSTITUENTS OF BODY AND FOOD. 



Globin prepared from haemoglobin is stated to be free from ash. It is 

 perhaps hardly correct to say that the ash is an impurity, because it is 

 extremely probable that in their native condition the actual proteid molecules 

 are combined more or less loosely with inorganic substances. 



The process of incinerating has its drawbacks in determining the amount 

 of ash in a proteid ; for in the heating, some of the sulphur of the proteid, and 

 when phosphorus is present the phosphorus also, will be oxidised and form 

 sulphuric and phosphoric acids respectively. H. Schulz l has recently shown 

 that sulphates are formed in tissues as a result of drying them at 110 C. ; this 

 would occur to a greater extent still at the temperature necessary for ignition. 



The sulphur in proteids is in the body normally burnt off as 

 sulphuric acid, which leaves the body in the urine as sulphates. The 

 ethereal hydrogen sulphates of the*urine originate in the intestine, as a 

 result of putrefactive changes in proteids, 2 and when putrefaction is 

 hindered by the administration of large doses of iodoform in dogs, these 

 products do not appear in the urine. 3 Krliger 4 has shown that a part 

 of the sulphur in proteids is present in a condition of stable combination, 

 a part loosely combined ; the latter is removed by boiling with alkalis, 

 the former is not ; the proportions of the two differ in different proteids. 

 Among tbe primary decomposition products of proteid, thio-acids, of 

 which thioglycollic acid is probably the most abundant, are obtained. 5 



From the elementary analyses which have been made of proteids, various 

 observers have attempted to construct an empirical formula for certain typical 

 proteids, egg-albumin being the one usually selected. Thus Lieberkiihn 

 assigned to albumin the formula C 72 H 112 N 18 22 S ; Loew Q gives the same 

 formula ; Harnack 7 gives C 204 H 322 N 52 b 66 S 2 ; Schiitzenberger, 8 C 240 H 392 N 6B 75 

 S 3 ; and there have been others. The great divergence between these numbers 

 requires no comment. 



Equally conflicting results have been obtained in attempts to ascertain the 

 molecular weight of albumin. Lieberkiihn, in 1852, attempted to establish it 

 by analysing the copper compound of egg-albumin ; more recently, Harnack 

 has done similar work. But very little importance can be attached to 

 such work at present, for Chittenden and Whitehouse 9 find there is no 

 definite copper albuminate, but that there are several in the mixture ; and 

 equally variable results are obtained with other metals both with egg albumin 

 and myosin. 



Such researches lead to the same conclusion as dialysis, namely, that 

 the molecules of proteid are extremely large, but leave us quite in the 

 dark as to their exact magnitude. 10 It is possible that in the future the 



1 Arch.f. d. ges. Physiol., Bonn, 1894, Bd. Ivi. S. 203. See also Halliburton and Brodie, 

 Journ. Physiol., Cambridge and London, 1894-95, vol. xvii. p. 154. 



2 Baumann, Ztschr. f. physioL Chem., Strassburg, Bd. x. S. 123. 



3 Morax, ibid., S. 318. See also more recently Nuttall and Thierfelder on "Animal 

 Life without Bacteria in the Alimentary Canal," ibid., vol. xxi. p. 109 ; xxii. p. 62. In tins 

 paper it is shown that healthy animal life is possible without micro-organisms in the 

 alimentary canal. 



4 Arch.f. d. yes. PhysioL, Bonn, 1888, Bd. xliii. S. 244. 



5 F. Suter, Ztschr. f. physiol. Chem., Strassburg, 1895, Bd. xx. S. 564 ; E. Baumann, 

 ibid., S. 583, and VirchoivsArchiv, 1894, Bd. cxxxviii. S. 560 ; E. Salkowski, ibid., S. 562. 



6 Loew and Bokorny, "Die chemische Kraftquelle im lebenden Protoplasma, " Munich, 

 1882. 



7 Ztschr. f. physiol. Chem., Strassburg, Bd. v. S. 207. 



8 Bull. Soc. chim., Paris, Ser. 5, tomes xxiii. and xxiv. See also Schmiedeberg, Arch. f. 

 exper. Path. u. PharmakoL, Leipzig, 1897, Bd. xxxix. S. 1. 



9 Stud. Lab. Physiol. Chem., New Haven, vol. ii. p. 95. 



10 The large size of the proteid molecule can be very strikingly demonstrated by the fact 



