MILK- SUPPLY OF LONDON WORKHOUSES. 39 



Before leaving the subject of the normal composition of 

 milk, it is right to refer to the laborious investigation by 

 Goppelsroder (vide " Yerhandlungen der Naturforschenden 

 Gessellschaft in Basel," 1866), which, at first sight, would 

 seem to be in opposition to the above. 



In reference to Goppelsroder 's paper, the remark should 

 first be made, that that chemist does not appear to deny that 

 the solid residue in the milk of a herd of cows keeps con- 

 stantly above the level just indicated. The point which he 

 insists upon is, that the milk of a single cow sometimes falls 

 in richness below the normal level, and observations are 

 cited in support of this statement. An examination of the 

 results given in his paper does not lead me to a similar 

 conclusion. In his paper I find many determinations of the 

 solids in milk believed to be unsophisticated. Only four out 

 of the entire number fall below 12 per cent. Now, it is 

 obvious that, however constant milk may happen to be as a 

 matter of fact, it must always be possible, by a sufficient 

 multiplication of analyses, to exhibit an analysis showing the 

 sample of milk as having a composition outside the normal 

 limit of variation. In other words, there is such a thing as 

 error of experiment, and the question to be asked respecting 

 Goppelsroder's four isolated cases of milk, showing less than 

 12 per cent, of solids, is whether this divergence from the 

 standard composition is real, or only an error of observation ? 

 Two of these instances occur in Table I., at the begin- 

 ning of the paper. Among eighteen samples of milk, 

 yielded by a single cow in eighteen consecutive days, he finds 

 one sample yielding 10.69 per cent, of solids, and another 

 yielding 11.41 per cent. In the same table, Goppelsroder 

 records the percentage of cream, and the specific gravity of 

 the milk before and after skimming. It is remarkable that 

 the two low percentages of solids are not accompanied by low 

 yields of cream or low specific gravities. The former of the 



