Chap. IV. TUBULARIA COUTHOUYI. 269 



those partially resorbed {Fig. 4, b c e) might readily be mistaken for young, budding 

 medusoids, did not the nearly obsolete chymiferous system, and the ragged loose- 

 ness of the cellular tissue, indicate their true condition. Those which have just 

 begun to wither may be recognized by their much diminished size {Fif/. 4, d), 

 shrunken proboscis (PI. XXI Y. Fig. 25, d), and the great thickness of their walls 

 {Fiffs. 25 and 25% a^) when compared to those of the medusoids in full vigor 

 (PI. XXVI. Fig. 3). 



Emhryohgij . Proles hjdroidca. — Tubularia shares with Parypha and Thamnocnidia 

 the remarkable property of reproducing the hydroid form without the intervention 

 of the egg phase. The genus Tubularia has recently received our special attention 

 in regard to this jwint, and we can safely say that it is not possible to find 

 any resemblance to an egg in the contents of the cavity of the medusoid buds. 

 If the egg is present it must be under the disguise of an unusual form. Can 

 it be i^ossible that the Purkinjean vesicle does not appear until the yolk masses 

 have separated from the germ-basis, at a time when they are so opaque as to 

 hinder all ordinary chances of a view of their interior ? We have investigated 

 these masses at this stage, but have not been able to see any indication of that 

 characteristic vesicle ; so that, if present, it must have been quite small. The 

 germ-basis (PI. XXIV. Fig. 8, /) occupies the cavity of the disk from a very early 

 stage, and originates in the same way as in Parypha and Thamnocnidia. When 

 the medusoid has reached about two thirds of its size, the germ-basis {Fig. 14, /), 

 which heretofore has been colorless, assumes a dingy yellow color, which, with 

 increasing age, grows darker, until the germ-masses begin to separate from it 

 {Fig. 15,//^). The manner in which these masses separate from the germ-basis 

 is altogether different from the regular process of self-division, as may be seen 

 from a sectional view {Figs. 15 and 17). At fii'st, the furrows are few {Fig. 15), 

 and. probably, always begin at the distal end of the basis; nor do they appear 

 to trend in any particular direction, in preference to another, but take their course 

 as often obliquely, as transversely or longitudinally with the proboscis. In time, 

 the whole basis becomes cleft, to at least half its depth {Fig. 17), by numerous 

 anastomosing furrows, and then it has all the appearance of a normally self-dividing, 

 single germ {Fig. IG). The number of the resultant segment-masses amounts to 

 at least fifteen or twenty, and, after the separation of these, a few more are pro- 

 duced by the residual basis; so that, in all probability, as many as thirty germs 

 are successiveh^ developed in one parent medusa. The earliest separated germ- 

 masses are the first to develop, but as they could not all grow, with like rapidity, 

 to their full size, within the restricted cavity of the disk, it becomes a necessity 

 which amounts to a law, that some one or two of them shall j^recede the others 

 {Fig. 18,/^/^), and as these come to maturity (PI. XXVI. Fig. 3), and leave the 



