378 DR ARTHUR T. MASTERMAN ON THE 



however, there are no more planes formed with definite relations to the axes of the egg, 

 but each octant commences to divide in the most irregular fashion. For some time 

 one can recognise traces of the former inequality in the cells of the lower pole (fig. 9). 

 but soon the lower cells become divided up in the same way as the upper into cells of 

 the same size and the morula is produced. This type, with small variations, must be 

 regarded as the normal, if such a term can be applied at all. It is that described and 

 figured externally by Sars in his Norwegian species. In a second common type the 

 segmentation commences at one pole with the formation of a cap of small cells. 

 Radiating outwards and downwards from this pole small cells are, as it were, cut out of 

 the yolk by the formation first of nuclei by mitotic division, and then of cell-walls 

 enclosing them (fig. 11). As a rule, these are formed most rapidly at the surface of 

 the egg, the cells gradually coining to envelop the central mass of yolk ; but as the 

 formation of cells always proceeds into the deeper parts before the whole yolk is 

 enveloped, and as the contained yolk does not differ in structure from the outer layer of 

 cells, but is destined very soon to be similarly segmented, we cannot compare this type 

 to the true superficial. This method of formation is often found succeeding the eight- 

 celled stage of the first type. Of other types we may note the uncommon one in which 

 the egg divides into two equal blastomeres, and the formation of small cells commences 

 at the two poles and spreads round therefrom. 



But in whatever manner the segmentation proceeds, the ultimate result is always a 

 solid morula (fig. 3) of very characteristic appearance. The cells are all of the same 

 size throughout and proceed to the very centre. Fig. 14 is a median section through 

 this stage. It will be seen that each cell has a nucleus at its centre, and the minute 

 spherules of yolk are arranged in a radiating manner, separated by fine strands of 

 cytoplasm (fig. 15). The cell- walls are clearly distinguishable, but the cytoplasm tends 

 to gather at the periphery of the cell, thus forming a definite meshwork throughout 

 the egg. On the average there are about twelve cells from the edge of the egg 

 to the centre. 



A fairly regular type of segmentation, with the production of a monoblastic 

 blastula, has been described in so many echinoderms of all five classes (the examples are 

 so numerous that it is needless to particularise), that Cribrella must in this particular be 

 regarded as a remarkable exception. Even so close an ally as Asterina has a perfectly 

 definite blastula stage with no morula (Ltjdwig, MacBride). At the same time it will 

 be evident later that Ophiura brevis'pina may possibly have a morula stage similar to 

 the above, and possibly also Solaster. Future work can alone determine these 

 points. 



Gastrulation. — The first question which would naturally occur to any observer of 

 this morula is — How is gastrulation to be effected ? The answer to this query has been 

 by no means easy to obtain, and has involved an amount of labour which is not 

 commensurate with the result. At the same time it has its points of interest. 

 Externally we notice the following changes. The morula gradually loses its reticulate 



