and two Esperiadge yrow the West Indies. 277 



was taken from a specimen equally small with the foregoing, 

 consists of: — (1) acuate fusiform, with headless wide than the 

 thickest part of the shaft (fig. 7, a) ; (2) bihamate contort, 

 thin, almost semicircular (c) ; (3) inequianchorate, large, three- 

 fluked («', b) ; (4) minute acuate in bundles {d) ; measuring 

 respectively 35-40-, 3-, 4-, and 5-6-1 SOOths of an inch 

 long. 



The bihamates and minute acuate spicules in bundles 

 occur together in masses in Esperia socialis. The latter, too 

 (viz. the " bundles," which do not appear to be initiatory 

 stages of the large spicules, but distinct developments, as I 

 have only found them in three sponges), are not less charac- 

 teristic of Stelletta lactea (Annals, Jan. 1871, pi. 4. fig. 22). 

 I saw them also in the mounted specimen of Esperia dia- 

 phana^ Sdt., from the Gulf of Florida, in the British Mu- 

 seum. Schmidt {pp. cit. t. iv. fig. 13) gives the ancho- 

 rate only, which is decidedly the largest on record, and in 

 which he has been able to illustrate the course of the central 

 canal. 



Beside the spicules above mentioned, each specimen was 

 pregnant with the usual rosettes found in the Esperiadse, 

 which, as those who have seen Dr. Bowerbank's excellent 

 figure 297 (/. c.) already know, consist individually of an 

 assemblage of the full-grown anchorates placed foot to foot in 

 a radiating globular form. Here certainly the anchorates can- 

 not be regarded as "retentive spicules," unless they are for 

 carrying out bits of sarcode for reproductive purposes in the 

 form of gemmules. We have yet to learn the office of these 

 beautiful and ornamental little bodies. 



With reference to the anchorates in detail, it will be ob- 

 served that they have respectively three flukes or arms (figs. 

 7 & 8), that the two lateral ones are winged on to the shaft 

 (e e), and that the central one is expanded into a petaloid 

 form (/), supported inferiorly by a falcate web-like sep- 

 tum which connects the median line of the middle fluke with 

 this end of the shaft [g). A similar condition exists in the 

 foot {h) ; but here the alae are united to the sides of the mid- 

 dle fluke, by which the space between the falcate septum and 

 the alge, on either side, is converted into holes like nostrils. 



I allude to this more particularly, because, in the lateral 

 view, it often appears as if the anchorate had but two flukes, 

 whereby it has as often, under misconception, been termed 

 " bidentate " (Bk. figs. 136 & 137, 1, c.) ; indeed in these two 

 figm-es there are, to me, evidently three flukes ; and, further, 

 I much question, if every kind of anchorate were minutely 

 examined on all sides or in all directions, whether any would 



