10 Mr. W. J. Sollas on Stauronema, a new 



fibres are not^ strictly speaking, confined to horizontal planes ; 

 for they curve upwards in gentle ai'cs so as to suggest that they 

 once bounded and corresponded with the rounded edge which 

 in all probability terminated the distal margin of the plate, in 

 the same way as a similar edge now limits its lateral margins. 

 The oscules and excurrent canals are arranged so regularly 

 in the plate that they do not disturb the regularity of the fore- 

 going arrangement to any great extent, though in their imme- 

 diate neighbourhood the sexradiate nodes become grouped 

 round the excurrent canal, so as to be subordinate to it rather 

 than to the general structure ; thus some of the nodal crosses 

 are turned round 45° out of their normal position, so as, in 

 joining Avith the others, to surround the circular canal with 

 continuous concentric fibres ; and, at the same time, the fibres 

 actually forming the walls of the canal are both bent and 

 thickened in order to bring about their complete adaptation to 

 its circumference. These facts may be seen in sections, but 

 better perhaps by etching the oscular surface with acid, when, 

 on the solution of the matrix, the oscular network stands 

 freely out in relief, and with its slightly expanded termination 

 resembles in miniature the mouth of a waste-paper basket ; 

 one can then see, by looking down into it, by reflected light, 

 ihe adaptation in the arrangement of the nodes and the bending 

 and thickening of the fibre, from which results a circular net- 

 work with circular fibres forming the walls. One will also 

 discover that the oscular fibres are beset with rather short 

 conical spines (PL III. fig. 1), which sometimes are simply 

 spinous outgrowths,' but frequently also the sixth arm of a 

 nodal radiation, which, instead of passing into the network 

 as usual, points freely into the excurrent canal, just as happens 

 in the canals of Aplirocallistes. In direction they usually 

 incline outwards and towards the centre of the excurrent canal, 

 but not always ; in exceptional cases they are turned inwards, 

 and then seem to be related to the fine canals which open in 

 the meshes of the oscular network, since they spring from the 

 sides of the fibre about such a space, and point into the excur- 

 rent canal. With this modification the rule here, then, as in 

 Ajihrocallistes^ seems to be that the spines always point in the 

 same direction as the outflowing current which at one time 

 passed by them. It is possible that this arrangement indicates 

 a defensive function for these spines ; but, as an explanation of 

 their position, one may recur to the fact that Carter has traced 

 the development of the spicule from its mother cell*, and 



* Ann. & Mag. Nat. Hist. 1874, vol. xiv. p. 97, pi. x. ; 1875, vol. xvi. 

 p. 11. 



