522 OR MASTERMAN ON THE FURTHER ANATOMY AND 



Such is the very peculiar way in which the arrangement described in the adult is 

 obtained, and the meaning of their method of origin formed an interesting puzzle. A 

 comparison with the mode of origin of the tentacles in Actinotrocha has given a solution 

 which has satisfied me, and may, I hope, appear plausible to others. 



In Actinotrocha the tentacles, when stretched above the head as in the adult, are 

 arranged, viewed from above, as in fig. 68a. The thick epithelium corresponding 

 to that of the oral grooves is on the inside of each tentacle. They arise in suc- 

 cession on each side on the dorsal surface of the collar region, immediately behind 

 the pre-oral hood. A stage with only six pairs is figured for simplicity. In 

 comparing the arrangement with, that of Cephalodiscus, we note that whereas, in 

 the former case, the tentacles move round ventrally to surround the mouth and 

 epistome, in the latter this is impossible, because the buccal shield (homologous 

 with the pre-oral hood or epistome) is used as a sucker. In this case, therefore, the 

 plumes, after travelling part of the way towards the ventral surface (fig. 68b), turn 

 backwards and move up again. To put the process more correctly, the actinotrochan- 

 like ancestor of Cephalodiscus adopted an adhesive function for the pre-oral hood, 

 immediately upon the assumption of a partially sedentary existence. A general 

 morphological rule for sedentary animals with tentacles is that these organs become 

 directed forwards in front of the head, so that in this case the three last pairs, unable to 

 move forwards in the ventral line, moved forward and dorsally upward to take up the 

 position found in Cephalodiscus. 



If the ontogeny of the plumes recapitulated the whole of phylogeny, the first three 

 pairs of plumes should migrate in procession from the point of origin of the first (which 

 corresponds exactly with that of the tentacles of Actinotrocha) down to nearly the 

 mid- ventral line, and then upwards and backwards (ab. fig. 68b). Such an apparently 

 useless migration would not be likely to occur, and we do not find that these plumes 

 migrate even as far as (cd. fig. 68b). The migration would be gradually shortened till 

 the present condition is arrived at, in which each plume moves directly from its point 

 of origin to its later position. At the same time, it is evident that the rotation through 

 180° of each plume results in the three plumes having their oral grooves facing away 

 from the buccal shield, just as would occur in the phyletic migration. On the other 

 hand, the last three pairs, occupying a position corresponding with that of the last three 

 pairs in Actinotrocha, arise in an equally direct and simple way, as in this larva, and 

 undergo no rotation. 



The Development of the Sub-Neural, Gland and the surrounding Organs. 



In the early buds, as already stated, the sub-neural gland by a movement inwards 

 comes to lie with its distal end in the sub-neural sinus, but as development proceed, 

 the ccelomic wall immediately ventral to the tip of the gland becomes thickened and 

 tucked in, invading the sub-neural sinus as a short diverticulum of the pre-oral coelorae. 



