Prof. W, Thomson on the " Vitreous " Sponges. 121 



kinds of these spicules are more or less completely fused to- 

 gether, forming a continuous anastomosing network. 



Two forms of free spicules are extremely abundant through- 

 out the group. The first are simply fusiform, frequently 

 slightly curved, and often enlarged and tuberculated or other- 

 wise armed or ornamented at one or both ends. These spicules 

 vary greatly in length — from '05 to '5 of an inch in the skele- 

 tons of Habrodictyon and of the sponge-mass of Dictyonema^ 

 where they are grouped in fascicles and make up the greater 

 part of the flexible network, to 3 or 4 inches in the silky fringe 

 at the base of Eujphctella, and to the enormous length of from 

 18 to 20 inches in the wonderful vertical wisp which is popu- 

 larly known as the "glass rope" oi Hyalonema. 



These spicules have all essentially the same structure ; they 

 consist of extremely thin concentric layers of silica separated 

 by films of sarcode, and are traversed throughout their entire 

 length by a delicate canal, occupied in the fresh state by a 

 sarcode axis. 



The second form is called by Dr. Bowerbank " cylindro- 

 rectangulated hexradiate " (British Sponges, vol. i. figs. 185, 

 186). It consists of a central shaft with the ends often spined 

 or tuberculated as in the fusiform spicules. Near the middle 

 of the shaft four secondary branches, at right angles to one 

 another, form a cross, the radii at right angles to the axis of 

 the shaft. The central canal is very distinct in the main axis, 

 and sends branches into the four diverging radii. Rarely only 

 two secondary branches are produced, but this is evidently by 

 suppression. These spicules are large, sometimes *05 inch in 

 length. They are scattered irregularly among the fusiform 

 spicules in Hahrodictyon^ Eujylectella^ and Hyalonema^ and are 

 sometimes aggregated in groups. 



The fusiform and the hexradiate spicules are modifications 

 of one type. About the middle of one of the fusiform spicules, 

 whether it be taken from the coil oi Hyalonema, from the fringe 

 of Euplectella, or from the general skeleton of any vitreous 

 Sponge, if we use sufficient care and a sufficient magnifying- 

 power, we can always detect one or two fine cross canals cut- 

 ting the axial canal at right angles. When the cross canals 

 have an appreciable length, two or four slight bulgings on the 

 outer surface of the needle indicate their position (PI. IV. 

 fig. 1 c). It is remarkable that this hexradiate type of spicule, 

 which is so abundant in the vitreous Sponges, is unknown in 

 any other order. 



In Hyalonema the hexradiate spicules of the second form are 

 usually if not always perfect and symmetrical. InHabrodictyon 

 they are very frequently distorted, the rays 



