CALCAREOUS SPONGES. 751 



(which appears to be due to change of position during individual 

 growth), affords a clear indication of what has really taken place. 

 Finally, in the most advanced types, such as Grantessa intusa7-ti- 

 citlata, we find the pseudosagittal subdernial spicules assuming 

 great dominance, almost to the exclusion of the typical articulate 

 tubar skeleton, so that we arrive at the so-called inarticulate 

 type. 



The development of these characteristic spicules appears to 

 antedate the ajjpearance of a definite dermal cortex, for we 

 find in Sycon ensiferum Dendy a similar canting of certain of 

 the distal tubar triradiates, which renders this species almost 

 indistinguishable from Grantessa. Indeed, it is this out-turning- 

 of one of the rays of the distal tubar triradiates that has, in our 

 opinion, led to the formation of a dermal cortex, probably by the 

 drawing out of the soft tissues of the sponge with the rays in 

 question. Thus the origin of the dermal cortex in this family 

 would be intimately connected with the developmei^t of these 

 subdermal pseudosagittal triradiates. In the Grantiidse, on the 

 other hand, the dermal coi-tex appears to have originated in the 

 development of tangentially placed triradiates in a previously 

 aspicular pore-bearing dermal membrane. 



These views undoubtedly tend to bridge over the gap between 

 the Sycettidse and the Heteropiidpe, and indeed the more pi-imitive 

 species of Grantessa are differentiated from Sycon and Grantia by 

 very slight characters, and difficult to separate from them, but 

 the rotation of the triradiates in question appears to have formed 

 the starting point of a new line of skeletal evolution which seems 

 to us to deserve recognition as marking a distinct family. 



We consider the views here put forward as to the origin of the 

 subdermal pseudosagittal spicules of the Heteropiidao to be more 

 in accordance with observed facts than those previously suggested 

 by one of us (Row 1909) in regard to the "subdermal secondary 

 sagittal triradiates " of Grantilla, which seem to be pseudosagittal 

 spicules really similar to those of Grantessa. 



We have changed the spelling of the name of the family from 

 Heteropidse to Heteropiidae, the latter being more in accordance 

 with the usual practice. 



Genus 18. Grantessa von Lendenfeld [1885 B] (emend.). 



Diagnosis. Canal system syconoid. No colossal longitudinally 

 placed oxea. 



For illustrations of this genus see von Lendenfeld [1885 B] 

 and Dendy [1893 A]. 



The tubar skeleton in this genus ranges from articulate, with 

 very numerous joints, as in Grantessa sacca, G. erinaceus, G. hirsuto- 

 and G. hispida, to inarticulate or nearly so, as in G. glabra and 



