[NTEKXAl. ANATOMY OF 8(TUILLA. 
417 
its posterior outer side and possessing a blind extremity which 
is somewhat dilated. This blind extremity lies just under 
the junction of the free carapace (F. C.) with the body and 
just behind the diverticulum of the gut [G. D.) which will 
subsequently form the liver. Text-fig. 1 is plano-stereo- 
scopic (combines in one plane parts which are only to be 
found in separate sections and which would only be visible 
collectively nnder the microscope if the section were of 
sufficient thickness to contain them all), and represents the 
gland as seen in sagittal sections of the larva. In a series of 
transverse sections of the larva, the gland was only included 
in eight successive sections, each 10 jli thick, and a.ppeared as 
a small tube cut transversely owing to the gland slanting 
posteriorly in position. The gland is, in all probability, 
functional at this early stage of development, since, with the 
exception of those situated at the extreme end of the diverti- 
culum and at the external opening, all the cells (in the 
middle part of the gland labelled M. Mx. Gl.) possess the 
striated border so characteristic of the excretor}" cells of the 
gland proper in the fully formed organ and contain granules 
in the cytoplasm. The fact (perhaps I should say probability, 
since the sections do not afford quite conclusive evidence) 
that the extreme cells of the terminal dilatation do not possess 
a striated border certainly favours the conclusion that this 
region becomes the end-sac of the fully formed gland. If 
this last statement be true (and there is but little doubt 
about it), then, in view of the obvious continuity of the wall 
of the gland diverticulum with the hypodermis at the external 
opening (see Text-figs. I and 8) and the absence of mesen- 
chymatous cells in the vicinity of the gland (the gland is 
apparently not invested by squamous epithelium at this 
stage), it seems extremely probable that the whole of the 
diverticulum (and therefore the whole of the fully formed 
gland, end-sac and all) is the product of an involution of the 
hypodermis, i.e. ectoderm. The probability of this conjec- 
ture is discussed below. 
The Gland in a 12 mm. Larva, Squill a sp. — The stage 
VOL. 59, PART 3. NEW SERIES. 27 
