178 Schrader. 
position near the tip of the band, but owing to a flexure of the latter, 
comes to lie in a sort of pocket (Fig. 30). Regardless of their first 
appearance, it can now be definitely established that at the extreme 
tip, as well as extending under the mass of clear cells, there is another 
distinet set of cells. With further growth and extension of the tip, 
these latter cells are carried along, but the mound of clear cells is left 
behind (Fig. 31), attached to one side of the germ band. Sections 
taken at different angles demonstrate .also that the mound of clear 
cells is now divided into two parts. 
The symbionts, which have entered into association with the giant 
cells during this time, remain perfectly stationary until the embryo 
revolves. It is a significant feature that at the completion of this 
movement, the bipartite mound of clear cells comes to lie close to the 
symbiotic mass, which is now on the interior of the embryo (Fig. 32). 
The vaginal invagination, which makes its appearance at this time, 
arises under or very close to the aggregation of clear cells. The 
direction of its growth is toward the symbiotic mass, and, in growing, 
the clear cells are carried along with it. Ensueing contact of sym- 
bionts and clear cell aggregates establishes the final relationship of 
the two. 
The foregoing account shows that the peculiar aggregation of clear 
cells represents nothing more nor less than the early germ cells. 
Definitely distinct at an early period of germ band growth, it retains 
its relative position until, when the S form of the embryo is attained, 
the tip of the germ band grows away from it. Both the germ cells 
and the symbionts are external to the curve of the embryo, and 
separated from each other by almost half the length of the egg. These 
relations are altered completely after the revolution of the embryo, for 
then, although the germ cells retain place on the band, they, as well 
as the symbionts, are on the inside of the curve of the embryo, and 
moreover, are removed from each other by only a short distance. 
Finally even this short distance is eliminated by the outgrowth of the 
vaginal invagination which pushes the germ cells into contact with the 
symbionts. 
SHINJTs confusion of the several types of cells is primarily due 
to his failure to work out the maturation divisions—a pardonable 
omission, since, as has been mentioned, he was concerned only with the 
general course of development. In regarding the giant cells as germ 
cells he must have neglected all of the definite cytological features 
