No. 3.] THE EMBRYOLOGY OF THE EARTHWORM. 445 
due to a similar shifting of the seat of mesoblast-formation with 
reference to the blastopore. 
It appears impossible at present to determine the primitive 
origin of the material now segregated in the primary mesoblasts, 
for the extreme condensation of development involved in their 
origin has completely masked the original mode of development. 
There is, however, nothing in the history of the teloblasts to 
preclude the hypothesis that the walls of the ccelomic cavities 
were originally formed as a series of gut-pouches, as in Am- 
plioxus. The primary mesoblasts lie at the extreme posterior 
limit of the entoblast, and it is not difficult to picture the process 
by which a series of gut-pouches, successively formed at the 
posterior part of the archenteron, might be crowded further and 
further back in development, until the present complete segre- 
gation of the mesoblast in a single pair of the cells was attained. 
(Cf. Hatschek, Zoologie, p. 76.) I will only add that if the 
mesoblastic somites should ultimately be shown to be homolo- 
gous with gut-pouches, and if my view of the homodynamy of 
the prostomial and trunk-cavities be correct, the prostomial 
cavity (z.e., the primary body-cavity of the Trochosphere), would 
be comparable with an unpaired anterior diverticulum of the 
alimentary canal (as in Guxzda and other Platyhelminthes), or 
with one of the median chambers of the Actinozoa. 
It is interesting to recall in this connection Bateson’s discovery 
that the prostomial mesoblast in Sa/anoglossus is actually derived 
from an unpaired median diverticulum of the archenteron ; and 
the same author has shown that this diverticulum corresponds 
closely with the anterior diverticulum in Amphiorus, though the 
latter structure divides into right and left chambers at the time 
of its separation from the archenteron. 
Bryn Mawp, Pa., June, 1889. 
C. METHODs. 
After testing many different hardening fluids, I have found none to compare 
with Perenyi’s fluid, which gives uniformly the best results, both for sections 
and for surface-views of all stages, and is far superior to picro-sulphuric acid 
or corrosive sublimate. Flemming’s mixture of osmic, chromic, and acetic 
acids gives very clear differentiation of the middle stratum of the germ-bands 
after staining with haematoxylin, but in most respects it is far inferior to Pe- 
renyi’s fluid. The embryos were left in the fluid from 15 to 60 minutes, 
