X PODAXONIA 383 



In any case, however, we fear that Masterman's hypothesis cannot 

 be upheld. Even if the " trunk " coelom does arise in the way which 

 he describes, this does not necessarily prove a close affinity between 

 Fhoronis and the Protochordata, because we have already seen that 

 the origin of the mesoderm from 4d in Annelidan eggs must be 

 regarded as a modification of such a mode of development, and 

 Erlanger has actually described the mesoderm as arising as a pouch 

 in Paludina. 



The fatal flaw in Masterman's theory is the absence of a prae-oral 

 coelom in the Actinotrocha larva, and though it is conceivable that 

 the notochord of Cephalodiscus should be represented by a paired 

 structure in Actinotrocha, yet the glandular pouches of Actinotrocha 

 have no resemblance to a notochord. The notochord in all Proto- 

 chordata, and in Vertebrata, is a modification of the endoderm into 

 a supporting tissue, by an increase in thickness of the cell-walls of 

 its component cells and the degeneracy of their contained protoplasm. 

 Such a change can be seen in the endoderm of the solid tentacles 

 of the hydroids of Tubularia, for instance, as compared with the 

 endoderm of the hollow tentacles of Hydra. But the mere fact that 

 the cells composing these glandular pockets in Actinotrocha contain 

 large vacuoles, does not create any special resemblance to a notochord. 



On the whole, the early development, so far as it is known, 

 creates the impression of being a modified form of the development 

 described for Fhascolosoma. In both forms the nephridia arise as 

 ectodermal pockets which, subsequently, after the metamorphosis 

 has been accomplished, acquire openings into the coelom. The ciliated 

 prae-oral lobe of Phoronis may be compared to the prototroch of 

 Fhascolosoma. In both forms the metatroch is prominent. 



After the Actinotrocha has led a free - swimming existence for 

 some time, and has increased in size, an invagination of the ectoderm 

 appears on the ventral surface, mid-way between mouth and anus. 

 This pouch increases in depth until it reaches the intestine of the 

 larva, to which it becomes adherent. The intestine increases greatly 

 in length and is thrown into several loops (Pig. 311) ; the pouch is 

 also thrown into folds as it grows longer. 



At length a critical point of growth is reached, at which meta- 

 morphosis suddenly supervenes. The ectodermic sac is everted and 

 forms a huge evagination which constitutes the main part of the 

 body of the " worm." As the intestine was attached to the apex of 

 this sac, when this is evaginated the intestine is drawn out in a U 

 shape. The ciliated tentacles of the post-oral band fall off, but 

 from their bases grow out stumps from which the adult tentacles are 

 later developed. The prae-oral lobe disappears, according to Caldwell 

 (1883) it is bodily amputated and falls into the gaping mouth and 

 is there digested. The whole metamorphosis occupies only a quarter 

 of an hour. We ourselves can testify that on one occasion we left 

 an advanced Actinotrocha in a watch-glass, left the room for a short 

 time, and on coming back found a young Fhoronis. 



