No. 3.] THE EMBRYOLOGY OF THE EARTHWORM. 431 
of the head-cavity is concerned, but the other points are of 
secondary importance. In Polygordius and the other larval 
types the head-cavity is manifestly derived from the blastoccel 
and is always unpaired. My observations on Lumbricus leave 
no doubt that the head-cavity is median and unpaired in the 
foetal types also, and except for its very small size has precisely 
the same morphological character as in the larval types. [This 
result is entirely opposed to the observations of Kleinenberg, — 
which have been especially emphasized by Balfour, — according 
to. which the head-cavity in Lusmbricus arises by the fusion of 
a pair of lateral cavities apparently homodynamous with ¢ruzk- 
cavities.| “The mesenchymatous character of the cephalic mes- 
oblast is obviously of secondary importance; for it has been 
shown that its cells not only have a common origin with those 
of the trunk-mesoblast, but also extend throughout the trunk- 
region (migratory mesoblast). Kleinenberg has shown, moreover, 
that in Lopadorhynchus and some other Polychzta the splanchnic 
mesoblast is formed by a process essentially identical with the 
formation of the cephalic mesoblast, — z.c. by migration of the 
cells through the blastoccel and their secondary arrangement in 
a continuous layer,—and he has thus also proved that the 
relations of the cephalic cavity to the blastoccel afford no ground 
for Hatschek’s distinction between primary and secondary body- 
cavity. I may add that the same conclusion is reached by the 
study of the foetal types (though on a very different basis from 
that adduced by Kleinenberg), for in these forms the blastoccel 
completely disappears, and the head-cavity is a new formation. 
In the larval types the coincidence of the head-cavity with the 
blastoccel is an incidental result of the circumstance that the 
blastocoel is large and persistent while the cephalic mesoblast 
is of late origin and is never sufficiently developed to fill the 
space between ectoblast and entoblast. This coincidence has 
no more significance than the coincidence of the blastoccel with 
the archenteric cavity which occurs among certain Amphibia.! 
The fourth statement requires careful consideration. As far 
as Lumbricus is concerned, it is untrue in both particulars ; for 
the halves of the cephalic ganglion arise at the anterior ends of 
neural rows, quite separate from each other, but in direct con- 
1 See O. Schultze, No. 4o. 
