A FRESH WATER SPONGE FROM SAHLG ISLAND. MACRAY. 321 



ranging from 1.5 to microns thick; the rotules being plane 

 disks less than "2 microns thick from the point where the shaft 

 begins to swell into them, and from 10 to 16 microns in diameter, 

 commonly near 12 microns, laciniately blunt-lobed around their 

 margins, the shaft occasionally extending 1 micron beyond the 

 disks, making the exterior of the rotule slightly umbonate. 



Long birotules : From 35 to 50 microns, commonly from 40 

 to 45, with usually a smooth, uniform shaft about "2 microns 

 thick ; the rotules generally of 3, to 4 or 5 rays which are slightly 

 incurved, the rotule ranging from 8 to 14 microns in diameter, 

 commonly from 10 to 11 microns. 



Larger skeleton spicules : Slightly curved, microspined or 

 rough, tapering gradually from the middle, then more rapidly 

 towards the ends; ranging from 150 to 260 microns in length, 

 commonly from 180 to 220; and from 3.5 to 5 microns in 

 breadth at the thickest part. 



Intermediate skeleton spicules: Generally smooth, and from 

 2 to 3 microns thick, and from 150 to 200 microns in length, 

 numerous and generally fascicled into strands which are often 

 connected transversely by the larger spicules. 



Smaller strand spicules and filament strands : Finer strands 

 than those referred to above, appearing as if made up of continu- 

 ous filaments instead of spicules: strands commonly from 10 to 

 15 microns across, made up of about 20 filaments or more, each 

 about one micron thick, where broker, across appearing as if 

 they were flexible to some extent, the ends of the filaments 

 showing a more or less distinct curvature. Under the micro- 

 scope they appear identical with the more slender spicules with 

 the exception that they appear to be continuous in the strand. 

 Examined with polarized light they are visible in the dark field, 

 as are also, more or less, the strands made up of the interme- 

 diate spicules while the spicules are cemented together, suggest- 

 ing a peculiar colloidal siliceous or a spongin cement. When 

 heated with nitric acid these filaments appear first to separate 

 and break into pieces, then partly at least to disappear. At the 

 earliest opportunity I purpose to examine the nature of these 

 PROC. & TRANS. N. S. INST. Sri., VOL X. TUANS.-U. 



