72 the president's address. 



the whqle spicule ; at the same time the rays may become spiny 

 or branched in a variety of ways, or even soldered together to 

 form a solid siliceous ball (Geodia). 



The sigmatose microscleres are more remarkable and more 

 constant in form. They are essentially linear spicules, and 

 appear to be derived from minute diactinal oxea. These may be 

 straight (" microxea," fig. 15) or bow-shaped (" toxa," figs. 16, 18a, 

 186), or their extremities may become bent over to form hooks 

 (" sigmata," figs. 17a, 176, 19). A very peculiar modification of 

 the sigmata is found in the " diancistra " (fig. 21), which often 

 resemble nothing so much as pocket-knives with the blades half 

 open. From the sigmata have also doubtless arisen the " chelae," * 

 characteristic of the family Desmacidonidae, and, in my opinion, 

 the most wonderful of all sponge spicules. Three different chelae 

 are shown in figs. 22 24a. 



A typical chela consists of a curved shaft, bearing a number, 

 commonly three, of recurved teeth, resembling the flukes of an 

 anchor, at each end. The flukes are sometimes expanded into 

 thin blades, and so also may be the shaft. Sometimes the flukes 

 at the two ends of the spicule are equal in size (" isochelae," 

 figs. 22, 22a), sometimes those at one end are larger than those 

 at the other (" anisochelae," figs. 23 24a), while in the genus 

 Melonanchora a very curious effect is produced by the meeting 

 and fusing of opposite flukes of an isochela at the equator of 

 the spicule. Minute differences in the form and number of the 

 flukes and the shape of the shaft appear to be constant, at any 

 rate within the limits of a species ; indeed, the very numerous 

 species of Desmacidonidae are to a large extent distinguished from 

 one another by these characteristics (compare figs. 23, 23a, and 

 24, 24a). 



The same constancy of form is to be observed in the sigmata, 

 although here there is less scope for specific differences. In 

 both cases the spicule, instead of remaining smooth, may become 

 more or less roughened by the development of minute projections. 

 This is shown, for example, in the sigmata of the genus Par- 

 esperella (fig. 20), where a row of small projections, like the teeth 



* It is perhaps unnecessary to discuss here the evidence for believing 

 that the chelae have arisen from sigmata. It is derived partly from the 

 development of the chelae themselves and partly from the occurrence of 

 intermediate forms. 



