60 PROF. F. J. BELL ON THE [Feb. 19, 



bear them are all distinct, the spines along the middle line of the 

 arms are beginning to grow up, but only one interradial spine is as 

 yet developed ; the spines on the ventral surface are now more 

 regularly arranged. 



In the third or largest specimen (y) all the spines — those on the 

 distal marginal plates, those on the middle Une of the arm, the 

 central portion of the disk, and those that lie between the inter- 

 radial lines — are all long, strong, sharp, and prominent. 



It is clear, therefore, that we have here to do with a condition 

 in which the spines increase in size and number during the growth 

 of their possessor; this is to be insisted on, inasmuch as it is not a 

 condition which always obtains. In some species of Linckia we 

 find that the spines diminish in size as the form grows larger ; but 

 in that case we also see that increase iu size is accompanied by 

 consolidation of the skeletal plates — or, where spines are wanting 

 and plates are not very strong, arms are often found to have been 

 brokeu off or injured. 



The study of individual development and the consideration that 

 the larger the form the greater its need of defence, lead to the 

 supposition that the least modified Oreaster will be found to be 

 one that is not specially spinose. This consideration gains in force 

 when we know that a form with feebly developed spines such as 

 O. nodulosus has in the most striking fashion the characters of the 

 younger repeated in the older individuals. 



In the arrangement, therefore, of the species of the genus, we 

 shall have to commence with those which have the spines least well 

 developed. Next to the non-spinous condition of such a form as 

 O. nodulosvs, we should probably place those in which a few spines 

 are developed at the proximal end of the arm to defend, so far as 

 may be, the more central, and thereby more precious, portion of the 

 internal organs. The next line of defence is probably that of the 

 free end of the arms, next the ventral plates, and lastly the dorsal 

 surface. 



On the other hand, we have in Oreaster occirlentalis an example 

 of a species in which the marginal spines disappear during growth. 



Not only have we evidence of this in Prof. Verrill's original 

 description of the species, but more convincing proofs are afforded 

 by the three specimens in the collection of the British Museum : 

 the smallest of these, one presented and named by Mr. Verrill, 

 never has more than five of the superomarginal plates on either side 

 of any arm without spines ; and a few are also to be found on the 

 infero-marginal spines. The smaller specimen described by Verrill 

 has " from one to four (upper plates) that bear small, short, stout, 

 bluntconical spines near the end of the rays ; " this is very much 

 the condition in which I find a specimen collected by Mr. Lock- 

 ington in San Francisco Bay. The larger specimen described by 

 Verrill had no spines on the superomarginal plates ; while a yet larger 

 specimen collected by Lockington has on the terminal plate of some of 

 the rays three minute processes, distinctly smaller than those of the 

 other specimen from the same collector. 



