were hard to remove, as the residue fairly floated with oil. Only by
long-continued application of heat, and in analysis III. over sulphuric
acid in vacuo, could a constant weight be obtained.

I would have used sand in the drying, or Baumhauer's method of fat
extraction, but for the small quantity of milk at my disposal and from
fear of loss of fat in the latter case.

The fat in III. was determined by extracting the dried residue and also
with 20 c. c. of milk by adding alkali and shaking with ether, removing
and evaporating the ether and weighing the fat.

As is shown in the table the sp. gr. is very low, though the solids and
solids not fat are great. The ash, casein, and sugar are in about the
usual proportion. The weight of casein, it is true, is but half that of
the sugar. The milk indeed shows an unusually great preponderance of the
non-nitrogenized elements, and this seems to correspond with the wants
of the animal, since fatty tissues are greatly developed in elephants.
According to Mr. Cross, who has had large experience with these animals,
they are fatter in the wild state than in bondage. These specimens must
appear as exceptional; they may be considered by some as "strippings;"
but as against such a view we have the recurrence in each sample of
the same characteristics in the milk and a near correspondence in the
composition. As may be seen from the subjoined analyses, given by v.
Gorup Besanez,[1] the milk belongs to the class of which woman's and
mare's milk are members, especially as regards the proportion of the
non-nitrogenized to the nitrogenized elements.

[Footnote 1: "Lehrhuch der Physiologischen Chemie," pp. 423 and 424.]