512 REVISION OF THE AXINELLID^., ii., 



the case of the digitate example. Where the reticulation is 

 apparent, the dermal pores are arranged in closely situated, oval 

 or rounded groups, or "pore-areas"_(P]. xl., figs.5, 6) measuring 

 up to 0-5 mm. in diameter, the pores themselves varying in 

 diameter from less than 20 to upwards of 80/*; within the pore- 

 areas, the dermal membrane is reduced, owing to the presence 

 of the pores, to a fine, lace like network. Where the dermal 

 membrane is apparently non-reticulate, this is due to the fact 

 that the pore-areas are much smaller and much more widely 

 separated. 



Skeleton. — The structure of the skeleton is such as would result 

 if the sponge had consisted, in the first place, of a number of 

 independent, simple or branched, digitifoim upgrowths, each 

 with its own separate skeleton, and if subsequently these indi- 

 vidual upgrowths, by lateral expansion and coalescence, had 

 grown together into a single mass,* and their skeletons become 

 more or less interunited : or, in other words, the skeleton is 

 resolvable into similarly constituted, simpler components, the 

 arrangement of which conforms to that of a system of ascending, 

 branched axes. In order to convey an idea of the general con- 

 formation of the skeleton, therefore, it will be sufiicient to de- 

 scribe the structure and mode of arrangement of the skeleton in 

 a single such component (as shown to best advantage in a digitate 

 process of the semi massive specimen), and to explain the manner 

 in which interunion is efiected between the skeletal fibres of 

 different components. 



In each simple digitation, the skeleton consists (PI. xxxi., fig.4; 

 PI. xxxii , fig.l): (i.) of stout multispicular main fibres radiating 

 outwards, almost invariably without branching, from the axis of 

 the process in a direction perpendicular or nearly perpendicular 

 thereto, and at a considerable distance (usually not less than 

 1 mm.) apart from one another; and (ii.) of very much slenderer 

 connecting fibres, most abundant towards the axial region of the 



■■' Tlie occurrence of pebbles and small patches of coarse sand liere and 

 there in tlie interior of all these specimens, more especiallj' towards theii- 

 base, lends colour to the view that the massive body of tiie sponge actuallj' 

 has been formed by the coalescence of originally separate digitations. 



