520 PKOF. GILBERT C. BOURNE ON FIVE 



Other observers have given more or less detailed accounts and figures o£ the 

 relations of the tentacles to the mesenteries in various species of Edicardsia ; 

 and, comparing their work with mj own, it seems that the mesenteries and 

 tentacles are formed in a regular and, on the whole, consistent sequence, which 

 I shall attempt to describe in the following pages. Unfortunately, very little 

 is known of the development of the Edwardsife, and in no case has the order 

 of appearance of the micromesenteries and tentacles been determined. The 

 youngest larvse described have the eight primaiy macromesenteries fully 

 developed. 



Probably the youngest larva that has been studied is the one described as 

 no. iv, of the larvae with eight mesenteries by E. van Beneden (3). This 

 larva was not identified as an Edicardsia^ but it has all the appearance of 

 belonging to that genus. It was ovoid in shape, without tentacles or buccal 

 cone, and had the eight characteristic " Edwardsian '■* mesenteries, each 

 provided with a large and prominent reniform muscle-banner. Boveri (5) 

 obtained larvoe of similar character ; in the youngest stage they were 

 spheroid;il and apparently without tentacles, but one of them was reared in 

 an aquarium for three months and at the end of that time was 1 cm. long 

 and showed Edvvardsian characters, agreeing in colour and form with 

 E. clajMredii, but had only eight tentacles. 



Meyer and Mobius (17) in 1863 obtained many specimens of E. duodeciui- 

 cirrata {= E. lutkeni) with from eight to eleven tentacles, but only figure 

 one with nine tentacles. From this figure it appears that the insertions of 

 the mesenteries on the peristomial disc were marked by radiating lines of 

 colour, as is usual in Edwardsise, and it is evident that the eight Edwardsian 

 mesenteries are present and that an additional mesentery has been formed, 

 probably in one of the lateral chambers. A single tentacle corresponds to 

 each mesenterial chamber. With these may be compared the parasitic 

 larva of Halcampa cliri/santhellum, fully described and carefully figured 

 by Haddon (13). This larva has six pairs of Hexactiuian mesenteries, 

 but only eight tentacles, and it is noteworthy that, with the exception 

 of the two corresponding to the directive endocoeles, all the tentacles are 

 exocoelic ; no tentacles are as yet formed in the dorso-lateral and ventro- 

 lateral endocoeles. 



Putting this information together, it is evident that in the larval 

 Edwards'ia and also in the larval Halcampa tiie first tentacles to appear — 

 which I shall henceforth call the primary tentacles — are formed as pro- 

 longations of the eight primary chambers into which the coelenteron is 

 divided by the eight so-called Edwardsian mesenteries. For convenience 

 of description, these eight chambers may be called megacoeles. It should 

 be noted that in Hexactinians two of these tentacles — namely, those of the 

 directive megacoeles — become endocoelic, six become exocoelic, and, as 

 Faurot (10) has shown in the case of Ilyanthus parthenopceus^ these six are 



