288 BULLETIN OF THE UNITED STATES' FISH COMMISSION. 



enteron or gut, as the latter is pushed inwards from behind, we are not 

 yet ready to assert, but such a mode of origin appears possible, if not 

 probable. By the end of the first twenty -four hours of development 

 the germinal disk measures almost a line across, as shown in Fig. 4, and 

 the part of it from Avliich the body of the embryo will be developed is 

 the widened portion of the blastodermic rim r, just below e. The colls 

 composing the disk at this stage are already too small to be successfully 

 represented in figures of the size we have adopted, consequently the 

 blastodermic rim and embryonic portion of it will hereafter be merely 

 more densely dotted. 



In Fig. 5 the disk or blastoderm is represented at thirty-one hours and 

 twenty minutes after the commencement of development ; it now meas- 

 ures about a tenth of an inch across, but is still extremely thin and has 

 apparently added nothing to its substance by an incorporation of any 

 of the imderlying yelk. The blastoderm is here again viewed somewhat 

 obliquely, in consequence of which the rudiment of the embryo c appeals 

 to have its head end inclined to the right hand. The embryonic rudi- 

 ment is relatively small, much more so than in other forms in the samie 

 stage of development. When the blastoderm is viewed from the edge 

 in the living state, as a transparent object lying at one side of the vitel- 

 lus, the segmentation cavity sc is found to be exceedingly shallow ver- 

 tically, but its lumen may still^be distinguished. The embryo, however, 

 is much more clearly marked than in Fig. 4; it is more prominent and is 

 rapidly growing in length from the rim. towards the center of the blas- 

 toderm. This brings us to the consideration of the growth in length of 

 the embryo from the edge of the blastoderm. I am inclined to believe 

 that the theory ]Hit forward by liallbur (Comparative Embryology, II, 

 254) must be accepted with considerable qualiticatiou, as stated by him 

 in the following language : " The growth in length takes place by a 

 process of intussusception, and, till there are formed the full number of 

 mesoblastic somites, it is effected, as in Chietopods, by the continual 

 addition of tresh somites between the last-formed soniite and the hind 

 end of the body." The only apparent exception to this rule is in Elecatc 

 canadus, where it aj^pears that the segmentation of the mesoblast on 

 either side of the neurala or spinal rervous cord is continued backwards 

 so as to involve the rim of the not yet closed blastoderm, and that the 

 somites of th# hind end of the body are formed by the coalescence of 

 the blastodermic rim in the median line continuous anteriorly with the 

 primitive groove. Should this be found to be the constant mode of de- 

 velopment in Eleeate^ it will be necessary to accept in 'part the view 

 urged by His and Ilauber. It is to be observed, however, that the 

 segmentation of the rim of the blastoderm in Elecate proceeds from 

 before backwards, and that while it extends beyond the posterior 

 extremity of the neural cord and notochord, the unusual segmentation 

 of the rim of the blastoderm behind the proper embryonic body into 

 muscular half segments may be a mere acceleration or hastening of the 



