138 E. W. MACBRIDE. 



on one side of the ovoid gland ; and this side, as shown by 

 comparison of vertical sections transverse to the axis of the 

 stone canal, is the right, if we suppose the animal to be placed 

 with its mouth down and the madreporite posterior. 



That the interpretation given to these rudiments is the right 

 one is convincingly shown by the next stage (figs. 9 «, b). 

 Here both axial and aboral sinuses are shut oflF from the coelom 

 and from one another. The Urkeimzellen are more distinct, 

 and one sees also smaller interstitial cells amongst them. In 

 this stage also I have been able to detect a trace of the charac- 

 teristic fibres of the adult gland. In fig. 9 b one sees an un- 

 mistakable rachis (compare fig. 11) constricted off from the 

 swollen ventral part of the ovoid gland ; and yet, traced forward 

 for a section or two, this ends in a similar knob of cells to that 

 represented in fig. 8 c. This will, I hope, convince every one 

 that the aboral sinus grows out from the involution at the base 

 of the ovoid gland. These intermediate stages are, however, ex- 

 ceedingly difficult to find; and in the majority of cases when 

 one seeks them one finds in embryos of the same size either 

 that the rachis is not formed, or that it is formed the whole way 

 round. I have, however, succeeded in getting about half a 

 dozen specimens with a rudimentary rachis, though this was in 

 all cases very short. 



I have said that the aboral sinus extends laterally as a 

 tubular involution, and of course the genital rachis is an out- 

 growth (as seen from figs. 9 a, b) of the similar part of the 

 ovoid gland. If we now compare fig. 10 (which is a section in 

 a plane somewhat oblique to the standard direction) we see 

 the perfect continuity of the primitive germ-cells in the 

 genital rachis with those in the ovoid gland. At the same time 

 we notice a very interesting change in their form. From being 

 spherical in the gland their bodies in the rachis become long 

 and fusiform, and overlap each other, and even show irregu- 

 larities of outline, which might be termed pseudopodia. Their 

 nuclei at the same time change from a spherical to an oval 

 shape. 



These facts, I think, strongly support Haraann's idea that 



