No. 2.] THE M ESQ BLAST-BANDS IN ANNELIDS. 207 



on Lumbricus^ — a somewhat unsatisfactory suggestion, which, 

 however, had the merit of emphasizing" the importance of a care- 

 ful study of the relations between the germ-bands and the blas- 

 topore in the Polychoeta. During the past two years I have 

 investigated several forms, with this end in view, in the hope of 

 finding facts that might bring into some relation the teloblastic 

 origin of the mosoblast (represented by Clcpsiiic, etc.) and the 

 delaminate mode of origin {LopadorJiyncJius), but until recently 

 the puzzle remained as great as ever. During the past summer, 

 however, I was fortunate enough, through the kindness of Dr. 

 E. A. Andrews, to procure very abundant material for the study 

 of the early stages of two species of Nereis {N. linibata, Ehlers, 

 and N. megalops, Verrill), and the facts thus brought to light 

 point the way, as I believe, to a solution of the problem. The 

 interest of the results depends largely upon the completeness 

 with which the early stages can be followed in detail ; and this 

 in turn is owing to the extraordinarily favorable character of the 

 eggs for observation. They are transparent, of comparatively 

 large size, and they may be procured in abundance. The four 

 primary entoblast cells {which are the remnants of the four 

 macromeres of the typical eight-cell stage) remain undivided 

 until a very late stage ; i.e. until the trochophore form is at- 

 tained and the eyes and mouth have appeared. Thus the axes 

 of the embryo may be located with the greatest precision 

 throughout the entire process of cleavage and gastrulation ; and 

 the certainty of the orientation is increased by the following 

 circumstance. The transparent macromeres contain large oil- 

 drops, which run together during the development, until, in the 

 great majority of cases, only four are left (one in each macro- 

 mere), viz. : a pair of smaller and a pair of larger drops, in corre- 

 spondence with the size of the macromeres in which they are 

 respectively contained. The smaller pair mark the anterior ex- 

 tremity, the larger the posterior, and they are bilaterally ar- 

 ranged on either side the middle line {cf. Fig. 6). There are 

 other advantages, equally great, that facilitate the precise study 

 of the early stages, but these will be described hereafter. 



Aided by these favorable conditions, I have been able to 

 trace the origin of the mesoblast-bands from the beginning of 

 development. As in LopadorJiyncJius and similar types, the 



^ Jour. Morphology, III., 1889. 



