of Cecidomyide Larve. 169 
Whether the appearance of a germinal membrane in our 
Cecidomyide larve is preceded by the separation of a structure- 
less peripheral layer, as has been proved by Weismann to be the 
case in the eggs of Chironomus and other Diptera, must be left 
undecided by me, from a deficiency of material for observation. 
Nor can I say how the germinal membrane is formed ; but there 
cannot be the least Hoan that it exists, and, as in true eggs, 
induces the series of embryonic developmental processes. 
The cells of which this germinal membrane is composed lie 
close together in a stratum, and have, as in Chironomus, an ex- 
tremely strong refractive power, so that it is difficult to detect a 
nucleus in their interior. At the posterior pole the cells are 
largest (0007 millim.), perhaps twice as large as at the oppo- 
site anterior end of the yelk—a difference which of course 
affects the thickness of the germinal membrane, move especially 
as from appearances it would seem that the posterior cells are 
arranged in a double layer (fig. 9). 
In somewhat larger germ-chambers (0°14 millim. in length, 
and 0:056 millim. in breadth) this difference is no longer per- 
ceived upon the yelk, which now measures 071] millim. Both 
before and behind, the cells now have exactly the same elongated 
form (fig. 10), and are of equal size; but even here the germinal 
membrane on the hinder part of the yelk is apparently composed 
of two superimposed layers of cells* 
The formative cells of the vitellus are reduced to three or four 
vesicular structures at the anterior pole of the yelk, although 
the form of the germ-chamber presents so far a certain amount 
of change, that the transverse section of the anterior seg- 
ment is but little less than that of the hinder one. In this 
condition I found the epithelial ling of the germ-chamber 
constantly converted into a granular layer. Whether this cha- 
racter is normal, I cannot say; but perhaps the circumstance 
that the later developmental stages of my larvee all died off and 
became transformed into a homogeneous granular mass might 
be connected with it. This supposition appears to me to be the 
more probable, as I some time since observed in the posterior 
half of the ovarium of a sterile queen bee, where the epi- 
thelium of the egg-chambers had undergone a perfectly similar 
change, that the ova were decomposed, and finally broke up into 
little fragments, instead of being further developed. 
In particular cases, however, the destruction of the contents 
seemed only to have commenced at a later period, as was shown 
not only by the greater size of the germ, but also by the cireum- 
* Tab. 36. fig. 31 of Wagner’s memoir must, I think, be referred to 
a chamber with a germina] membrane; only it appears as if the formative 
cells of the vitellus had already completely disappeared. 
