No. 3.] THE EMBRYOLOGY OF THE EARTHWORM. 391 
arises by the forward growth and union in the median line of 
the mesoblastic bands, a process which is effected by prolifer- 
ation and migration of the mesoblast cells already formed, and 
not by the formation of new mesoblastic elements from the 
superjacent ectoblast. The prostomial cavity is from the very 
first unpaired. 
(8) The przoral or cephalic ganglia are differentiated out of 
the front ends of the neural rows at a time when these rows are 
fused with the ectoblast, and before they have met in the 
median line in front of the mouth. There is therefore no 
median apical plate (Scheitelplatte), but only a pair of lateral 
ectoblastic thickenings continuous behind with the neural rows. 
I am unable to say positively whether the cephalic ganglia actu- 
ally arise from the neuroblasts, but I believe they do not [p. 416]. 
For more special accounts of the origin and differentiation of 
the layers and the detailed development of organs, the reader is 
referred to Part II. It is, however, desirable to preface the 
special descriptions with a short discussion of their relation to 
the work of other investigators. 
The principal questions suggested by my work relate to the 
origin of the mesoblast and the history of the middle stratum of 
the germ-bands. The existence of this stratum in the germ- 
bands of Oligochzeta has not hitherto been distinctly recognized, 
though Kowalevsky and Kleinenberg indicate it in their figures 
of Lumbricus. The nephric and outer cell-rows and the four 
anterior teloblasts have hitherto been observed only in the 
Hirudinea and adequately studied only in the single case of Clep- 
sine. As regards the mesoblast, I am fully convinced that the 
whole of the “epiblastic mesoblast”” which Kleinenberg believed 
to be derived directly from the outer layer, is simply the middle 
stratum which he failed to distinguish from the inner or meso- 
blastic stratum. This misinterpretation has importance because 
of the emphasis which Balfour placed upon Kleinenberg’s con- 
clusions, and still more on account of its connection with the 
views put forward in Kleinenberg’s recent remarkable paper on 
Lopadorhynchus (No. 31). In this masterly work the author not 
only rejects the ordinary conception of the mesoblast, but even 
goes so far as to deny its existence as a primary feature of devel- 
opment, all “mesoblastic”’ structures being conceived as direct 
or indirect derivatives of the two primary layers, ectoblast or 
