30 ORGANIC SUBSTRATA OF THE ANIMAL ORGANISM. 



still retaining water. Hence it is evident that the number of atoms 

 of hydrogen will be computed with the least certainty in the most 

 important elements of zoo- chemistry, as in the albuminous matters 

 arid their derivatives, which are bodies of very high atomic weight. 



In consequence of the atomic weights of these substances being 

 so high, and considering the great uncertainty whether they are free 

 from all admixtures, excepting the salts with which they are insepa- 

 rably connected, the number of atoms of carbon cannot be computed 

 with certainty from the empirical result of the analysis. As, more- 

 over, we possess no means of directly determining the oxygen con- 

 tained in an organic body, and can only estimate it by the loss in 

 weight of the substance analysed, that is to say, by the subtraction 

 of the quantities of carbon, hydrogen, and nitrogen, the collective 

 errors in the investigation will frequently affect the number repre- 

 senting the oxygen, which must therefore be regarded as the most 

 uncertain number in the analysis. 



When all the errors which attach to the calculation of atomic 

 formulae from the direct results of elementary analyses have 

 been as thoroughly as possible avoided, and even when they 

 may be regarded as = 0, the formula will still only have a 

 problematic value until the saturating capacity of the body has 

 been determined by direct experiment, that is to say, until the 

 atomic weight derived from the saturating capacity of the body shall 

 be found to accord with that deduced from the analysis. We have 

 therefore no guarantee for the true atomic weight of a body, or 

 for its atomic composition, without a previous knowledge of the 

 saturating capacity, even supposing that all the other data were per- 

 fectly correct, and free from doubt. Thus, for instance, we should 

 not know whether lactic acid and starch were composed according 

 to the formula C 6 H 5 O 5 , or C 12 H 10 O 10 , or according to other 

 multiples. But there are, unfortunately, many animal substances 

 of a higher order, whose atomic composition cannot be tested by 

 a comparison with their saturating capacity. Such substances either 

 do not combine in definite proportions with other substances, or 

 do so in various relations, so that it is impossible to determine 

 which combination is actually to be regarded as the neutral one. 

 The variations in the numbers of the saturating capacity, are fre- 

 quently much more important in such bodies (partly owing to the 

 admixture of mineral substances with them) than those of the num- 

 bers of the elementary analysis ; that is to say, the atomic weight 

 derived from the saturating capacity is frequently no less uncertain 

 than that derived from the elementary analysis. 



