630 



SCIENCE. 



[N. S. Vol. V. No. 120. 



first bilaterally symmetrical division in the egg 

 (58-cell stage) is the fourth division of the blasto- 

 meres of the first quartet, viz., the formation 

 of the apical cross of eight cells. There are no 

 head-kidneys, and the cell corresponding to the 

 nephroblast remains on the surface as a small, 

 inconspicuous ectodermal cell. 



The ventral or somatic plate grows as in 

 Amphitrite. The second bilaterally symmetrical 

 division in the egg is the third division of the 

 large posterior cell (X). A smaller cell is given 

 oflF anteriorly and lies across the median plane. 

 The somatic plate broadens laterally and pos- 

 teriorly, the lateral portions growing posteriorly 

 faster than does the central portion, so that 

 finally the cells along the posterior edge begin to 

 meet in the median line just dorsal to the blasto- 

 pore. The point where this concrescence begins 

 is the point where the paratroch appears. This 

 organ is formed from derivatives of X, but by 

 an entirely different set of divisions from those 

 which take place in Amphitrite^ according to 

 Dr. Mead's very kind personal communication. 

 Within the paratroch are enclosed certain small 

 cells which later become the proctodseum. 



The first bilaterally symmetrical division of 

 the mesoblast occurs at the seventy-cell stage. In 

 the entomeres a fifth quartet is formed and the 

 three cells of the fourth quartet divide bilater- 

 ally. No further divisions occur before the 

 closure of the blastopore. 



The first bilateral division, in the small ecto- 

 meres, in the large X cell, in the mesoblast and 

 in the entomeres all occur at very different 

 periods in development but in the same generation 

 of cells, viz., the eighth, counting the unseg- 

 mented egg as the first. 



Gastrulation is a combination of invagination 

 and epiboly. The mesoblast shows the first 

 sign of invagination; then the entomeres also 

 elongate inward until they form a column ex- 

 tending to the ectoderm of the upper pole. 

 Now the ectomeres overgrow their lower sur- 

 face from the sides and ventrally, forming a 

 triangular blastopore with its short base directed 

 ventrally. The cells forming the lateral and 

 ventral lips of the blastopore are twelve deriva- 

 tives of the third quartet of ectomeres and later 

 form the stomodseum. The proctodseal region 

 should probably be regarded as a part of the 



blastopore, for it is originally continuous with 

 the rest of the opening. It becomes separated 

 from the stomodseum by the postero-anterior 

 concrescence of the somatic plate. The gastrula 

 is bilaterally symmetrical and its axis corre- 

 sponds to none of the principal axes of the 

 adult. 



During the closure of the blastopore the meso- 

 blast bands are forming in the interior of the 

 egg. In this case we have an actual change of 

 the plane of the mesoblast bands through nearly 

 90°, i. e. , from nearly dorso-ventral to longi- 

 tudinal. The first cells of the bands are given 

 off almost ventrally, but with each successive 

 division of the mesoblasts the direction of the 

 spindle is more nearly longitudinal with respect 

 to the egg, until finally the cells of the bands 

 arise anteriorly from the mesoblasts. 



A word concerning the cleavage of Sternaspis 

 scutata may be added here. This was worked 

 out as far as the 80-cell stage by the author of 

 this paper and corresponds, cell for cell, with 

 that of Arenicola, differing, however, as regards 

 size and structure of blastomeres and order of 

 their formation. Sixteen relatively large cells 

 corresponding to the primary trochoblasts of 

 Arenicola are formed, but Sternaspis possesses 

 no prototroch and these cells form part of the 

 ectoderm. 



II. The Oblique Cleavage and its Relation to the 

 Mosaic Theory. After a review of the positions 

 held by Roux, E. B. Wilson, Driesch, Hertwig, 

 etc., the following arguments were given 

 against the mosaic theory as applied to the 

 oblique cleavage. 



1. The failure of cell-homology in a rapidly 

 increasing number of cases. 



2. The very different size and structure in 

 different species of blastomores which have the 

 same normal fate. 



3. Experimental work. Professor Wilson and 

 Mr. Crampton err in regarding the experiments 

 of the latter on Ilyanassa as supporting the 

 mosaic theory, for in most cases ' regeneration ' 

 occurs. Mr. Crampton himself states in the 

 text that in I and \ embryos the endoderm cells 

 are completely overgrown by ectoderm, and also 

 gives figures illustrating the same fact. It is, of 

 course, impossible for ectoderm cells, which 

 normally cover only the outer surface of cer- 



