178 ROBERTS, ON BLOOD-CORPUSCLES. 
of a similar nature, but modified in some very peculiar man- 
ner. Its first operation is to cause the corpuscle to en- 
large by imbibition, and this goes on progressively until at 
length the cell is destroyed. If the solution be strong, this 
destruction supervenes at once. The tannin also unites with 
the cell-contents and coagulates them, imparting to the cor- 
puscle, finally, a solid consistence. ‘The conditions of the 
imbibition are disturbed by the previous application of ma- 
genta; for no pullulation, or at most only traces, occurs when 
the corpuscles are treated first with magenta and then with 
tannin. 
The bearing of these observations on the current views 
respecting the structure of the vertebrate blood-dise is im- 
portant. They seem to warrant the inferences drawn in the 
two following paragraphs: 
1. The exact identity of the appearances produced in the 
blood-discs of the ovipara with those observed in the mam- 
malian corpuscles lends strong support to the view that these 
corpuscles are homologous as wholes; and that the mam- 
malian blood-dise is not the homologue of the nucleus of the 
coloured corpuscle of the ovipara, as was conceived by Mr. 
Wharton Jones. 
2. The observations likewise lead to the behef that the 
envelope of the vertebrate blood-disc is a duplicate mem- 
brane; in other words, that within the outer covering there 
exists an interior vesicle, which encloses the coloured con- 
tents, and, in the ovipara, the nucleus. 
Dr. Hensen,* of Kiel, had already, in 1861, convinced him- 
self, from wholly different observations, that the blood- 
corpuscles of the frog possess such a structure. On this 
view the blood-corpuscle is anatomically analogous to a 
vegetable cell, and the inner vesicle corresponds to the 
primordial utricle. 
The present observations indicate, by direct proof, a dupli- 
cation at only one or, at most, two points in the blood-dises 
of mammals and birds. Nevertheless certain appearances, 
occasionally observed, favour the notion of a complete dupli- 
cation. (Fig. 1, 0.) 
The admission of this hypothesis, however, scarcely re- 
moves the difficulties sufficiently to permit a tenable expla- 
nation to be offered of the appearances described in this 
paper. Yet, as it may prove suggestive to some other 
inguirer, [ will not suppress what appears to me the expla- 
nation least open to objections. It might be conceived that 
the cells enlarged by imbibition, until at length the less dis- 
* ¢Zcitschrift fiir wissench, Zoologie,’ Band xi, p. 263. 
