36 THE president's address. 



directed vertically inwards towards the gastral cortex and in- 

 creases greatly in length. 



We thus have a complete reorientation of the spicule in ques- 

 tion, and a new symmetry established as regards its rays. We 

 have what we term a pseudosagittal spicule, superficially re- 

 sembling the ordinary sagittal spicule, but in which the indi- 

 vidual rays do not correspond as they appear to with those of 

 the latter. The apparently unpaired and very strongly de- 

 veloped ray (which we may now call the centripetal ray) is one 

 of the original paired (oral) rays, and one of the apparently 

 paired rays (which we may now call the outer or dermal rays) 

 is the original unpaired (basal) ray. The new symmetry never, 

 so far as we have observed, becomes perfect, for the outer or 

 dermal rays usually differ from one another in shape. Some- 

 times one is straight and the other curved, and the straight one 

 represents the original basal ray. The centripetal ray grows 

 much longer than the dermal rays and may even reach the 

 gastral cortex. 



In the meantime the sagittal triradiates of the chamber 

 skeleton which lie nearest to the gastral cortex have become 

 specialised as subgastral sagittal triradiates. This change 

 involves little, if anything, more than increase in length of the 

 centrifugally directed basal rays, which interlock, as already 

 explained, with the centripetal rays of the subdermal pseudo- 

 sagittals. By disappearance of the remaining spicules of the 

 original articulate tubar skeleton the transition to the " inarticu- 

 late " type is completed. 



The second way in which an " inarticulate " type of skeleton 

 may arise is seen in the Amphoriscidae. Here the subdermal 

 spicules are quadriradiates, the large centripetal ray being an 

 apical ray added to an ordinary triradiate whose three facial 

 rays lie tangentially in the dermal cortex. Obviously such a 

 skeleton could not arise until the dermal cortex had become 

 well established. Here, then, we have a beautiful example of 

 convergent evolution, the same mechanical result being broughr 

 about in two totally different ways in the two families Heter/ 

 piidae and Amphoriscidae. 



It is extremely interesting to observe that after the skeleton 



