SEMAROSTOME^ — AURELLIA. 625 



convoluted ridges on the subumbiella Hoor ot" the 4 genital cavities. The liases of the genital 

 ridges are beset with numerous, small, gastric cirri. 



The gelatinous substance of the disk is of a translucent milky-white or yellowish-brown; 

 spermaries usually slightl)- pink. In old individuals the gonads in both sexes are white. 



Common from Greenland to the West Indies. At Eastport, Maine, it is mature in Sep- 

 tember, and at Tortugas, Florida, in May. 



The American medusa is closely allied to Aurcllia aurita of Europe and is at most 

 merely a variety of the latter. It may possibly differ from its European representative in 

 the thickness and rigidity of the mouth-arms, which are very broad at their bases and often 

 complexly convoluted at their tree edges, but I have seen these same characters in /luttllin at 

 Naples and consider the American and European medusa to be identical. 



A very complete description and numerous figures of the American medusa are given 

 by L. Agassiz, 1860-62, Cont. Nat. Hist. U. S., vols. 3 and 4. 



Development. — The ova are dehisced from the gonads into the interradial grooves from 

 which thev enter the median gutter of the mouth-arms and are here retained in small pouches 

 near margins of free edges of mouth-arms and finally set free in the planula stage. Minchin, 

 1889, gives a good description of these brood-pouches. Segmentation total and unequal, and a 

 blastula is formed which has a large, central blastoccel. According to Hyde, 1894, the gastrula 

 may be formed in either one of two different wa)s: (1) bv the invagination of a small part of 

 the blastula wall combined with the ingression of numerous cells from various parts of the wall 

 of the blastula; (2) by invagination of the wall of the blastula, aided only occasionally by the 

 ingression of cells from the blastula wall. According to Smith, 1891, however, the gastrula is 

 formed from a small invaginated region in the wall of the blastula, from which there develops 

 a single, continuous layer of cells, which layer finally completely fills the cleavage cavity, thus 

 giving rise to a 2-layered embryo with an open blastopore. Smith denies that this process is 

 aided in the least by the ingression of cells from the wall of the blastula into the blastocoel. He 

 finds, indeed, that a few cells are occasionally seen to wander into the blastula cavity, but these 

 always degenerate without taking an\' share in the formation of the entoderm. These varia- 

 tions in the mode of forming the gastrula have been seen in other Scyphomedusse, having been 

 observed by Conklin in Linuche, and Hyde and McMurrich in Cyanea. The blastopore 

 then closes and the entoderm becomes a closed sac entirel)- enveloped b\ the ectoderm. The 

 larva then becomes ciliated and swims actively about as a pear-shaped planula, which soon 

 attaches itself to the bottom by the wide anterior end. A crater-like depression (oesophagus) 

 formed of ectodermal cells then appears at the narrow (now the upper) end of the animal, and 

 this presses down upon the entodermal sac. The first pair of radial stomach-pouches is formed 

 from the entodermal sac, while the second pair is formed, at least partially, from the ectoderm 

 of the cup-like depression. The mouth breaks through and 4 tentacles appear. The larva 

 then has 4 interradial, longitudinal, partial septa each formed of a fold of the entoderm sup- 

 poned by a central shelf of gelatinous substance. These septa extend from the margin of the 

 mouth to the lower end of the stomach-cavity. They form the 4 primary, gastric filaments 

 of the future eph\ra, and there are no septa in the central stomach of the medusa. 



As we have stated, it appears from the researches of Gotte and of H\'de that the two 

 original pairs of stomach-pouches are derived alternately from the ectoderm of the oesophagus 

 and from the entoderm of the primitive stomach, although Hyde shows that the lower aboral 

 floor of the 2 oesophagus pouches is formed at least partially from entoderm. Through 

 division of the 4 original stomach-pouches we have finally 24 pouches, 10 entodermal and 14 

 mainly ectodermal, as follows: 6 diammetrically opposite perradial, 4 interradial, and the 4 

 connecting adradial pouches are ectodermal. 90° apart from these the 6 perradial and their 4 

 adjacent adradial pouches are entodermal. (See Gotte, 1887.) 



One must remember that R. P. Bigelow, 1900, finds that the 4 piimar\- stomach-pouches 

 oi Cassiopea xamachatia are wholly entodermal, and Hadzi, 1907, finds that this is also the case 

 in Chr\'saora. Moreover, according to Hadzi there is no ectodermal invagination in Clirysaora 

 to form the mouth, but on the contrary the throat is evaginated and lined with entoderm. 



The larva of A urellia becomes a scyphostoma which finally attains a height of about 5 mm. 

 and acquires 8, 16, and finall\- 24 long tentacles. The eph\ i.t are developed through strobili- 

 zation of the scyphostoma. As man\- as 13 annular constrictions ma\- develop below the zone of 



