156 Records of the Indian Museum. [Voi,. X, 



their extremities, or at an^^ rate at the extremity nearest the 

 surface. They are usually somewhat contorted, but they do not 

 anastomose; their course is mainly horizontal but as a rule dips 

 down into the sponge more or less deeply ; they are connected 

 together to form a somewhat loose reticulation by single tornote 

 spicules. The styli and tylostyl form in the parenchyma a fairly 

 regular reticulation for the most part composed of single spicules 

 and comparable with the t3^pical skeleton of Reniera; but traces 

 of fasciculation can be detected at some points. 



Spicules: Macrosclcres. — The majority of the dermal macros- 

 cleres are tornota with well-defined smooth oval extremities of 

 comparatively large size. Their shafts are as a rule smooth, 

 slender and straight. Both amphioxi and amphistrongyli occur, 

 however, among them, always sparingly. These spicules, which 

 must be regarded as abnormalities, invariably have their shafts 

 irregular in outline and as a rule are inflated at several or many 

 points. The skeleton-spicules are as a rule a little shorter and 

 stouter than the tornota; their shafts are usually smooth, but 

 occasionally bear a few scattered spines near the blunt end ; this 

 end, which is in most cases distinctly globular and well differentia- 

 ted from the shaft, is rarely or never quite smooth, but as a rule 

 may be called irregular in outline rather than actually spiny. The 

 other extremity is sharply and gradually pointed. These spicules 

 are from 0124 mm. to 0"i6 mm. long. Sometimes much shorter 

 and stouter tylostyles of very irregular form occur in small numbers, 

 and even what may be called normal spicules of the type vary 

 considerably both in proportions and in outline. Typical tornota 

 are on an average about 0"i67 mm. in length (o't66 to o*i76 

 mm.), 



Microscleres. — The microscleres include minute and very 

 slender C- and S-shaped sigmata as well as isanchorae. In the 

 latter the shaft is stout, somewhat compressed laterally and 

 feebly curved ; the three teeth at either end are subequal, narrow 

 and sharply pointed ; those at the sides project outwards in such 

 a way that it is hardly possible for the spicule to rest on its 

 dorsal surface. (This makes it impossible to obtain an accurate 

 camera lucida drawing of the front view). 



Type No. Z.E.V. ^7" Ind. Mus. 



Locality. — Madras harbour in 4 or 5 feet of water ; on living 

 shells of Mytilus latus, I^am. (together with Balanus amphitrite, 

 Darwin) and also on those of Ostrea sp. 



This sponge would not fall within the genus Lissodendoryx 

 as originally defined by Topsent, for all the parenchymal 

 macroscleres are not smooth , although most of them are nearly 

 so. Lundbeck has, however, pointed out that the critical charac- 

 ter lies not in the form of the macroscleres, but in that of the 

 hooded microscleres. These belong to the type known to him and 

 to some other writers as iso-anchorae, whereas the corresponding 

 spicules in Myxilla are true iso-chelae. The distinction may be 

 accepted as convenient; but it should be noted that the impor- 



