374 MR. J. D. MACDONALD ON THE ANATOMICAL CHARACTERS 



investing membrane of which the tail received a covering. These cells no doubt form 

 the proper test of the embryo, and as such must also be the basis of the connecting sub- 

 stance or common test of the mass. 



The part destined to become the future abdomen was divided into several rounded 

 lobes, from the confluence of which the taU emerged posteriorly, and two distinct bran- 

 cliial chambers arose in front. 



A nervous ganglion with an otolithic sac, and what may probably be a rudimentary 

 ^"isual organ, Avere connected with one of the thoraces ; while the three sucker-tubes and 

 foiu- glandular stolons, above referred to, originated near the pedicle of the other. 



The incipient branchial network was marked by several transverse rows of short slits, 

 which appeared to have thickened or everted edges, on account of the highly-refracting 

 property of the epithelium at this stage of development. 



There are many interesting questions connected with this double form of existence, to 

 be answered perhaps by further observation of its structure, evolution, and history. 



The imusually large ova, which undergo the ordinary round of changes subsequently to 

 theu" liberation from the body of the parent, and are perfectly enclosed in the connect- 

 ing substance, seem to negative the idea of their being discharged in the usual way, and 

 rather suggest their escape by rupture of the abdominal walls, or by pedicellation and 

 ultimate separation from the abdomen. With this consideration before us, the query 

 naturally arises, Does the embryo at any stage become free and locomotive, as in other 

 cases ? and if not, which I am much disposed to believe is the fact, what purposes can the 

 transitory suckers, organs of sense, and taU subserve ? 



Thei'e being two respiratory chambers, with their proper openings, two mouths and 

 two vents, with a visceral mass common to both, forming perhaps the most perfect 

 diplozoon in creation, — now that the originally simple nature of Nordmann's celebrated 

 example has been satisfactorily explained, — are we to recognize here two distinct entities ? 

 If so, are the sexes and generative function common to both, divided between them, or 

 restricted to either ? Some species of Polyzoa clearly exhibit the union of two zooids 

 ab ongine in the formation of the so-called " ovicell" and its contents, which are very pro- 

 bably developed at the close of a definite cycle of gemmations. These ideas may be con- 

 sidered transcendental and visionary ; but surely the scope of our philosophy is legitimate 

 so long as oui- judgment in relation to immaterial and abstract things is circumscribed 

 by that evidence alone which material things afford. Be it as it may, however, each 

 little animal in the present genus, just as in the case of simple Ascidians, propagates its 

 kind both by gemmation and true oviparous development ; and the offspring appears to 

 differ in nowise, either in its structure or history, from the parent. No vascular bond of 

 union pervades the mass ; and but for the continmty of the cell-structure of the connect- 

 ing substance, it would be difficult to regard it as any more than a simple aggregation of 

 individuals. Indeed it is common enough to find the members of aggregate species of 

 simple Tunicata blended by the coalescence of their test or outer covering ; yet this 

 circumstance alone cannot be considered a valid reason for classifying such with the 

 compound forms. 



What then are the most important distinguishing features of the latter ? I think it 



