84 BULLETIN 82, UNITED STATES NATIONAL MUSEUM. 



whether there are always five basals, or whether the basals agree in number with 

 the radials. 



Among the remaining recent stalked crinoitls specimens with more than five 

 rays are practically unltnown, though I have at hand an example of Metacrinus 

 rotundus with eight rays and six basals. 



A number of 4-raj^ed .specimens have been reported in widely separated coma- 

 tulid groups. There is no suggestion that their occurrence is anything but for- 

 tuitous. In all cases where the missing ray has been identified it is the anterior. 



Only in the littoral fauna of the warm seas do highly multibrachiate species 

 with more than 40 arms occur, and the great majority of all the species with 

 more than 10 arms are found only in tropical and subtropical regions within 

 100 fathoms of the surface. A number of the genera characteristic of intermediate 

 depths in water of moderate temperatures belonging to the families Calometridse, 

 Thalassometridse, and Charitometridaj include species with between 15 and 30 arms; 

 but in the deep-sea and in cold water species with more than 10 arms are very rare, 

 and specimens with from 11 to 13 arms are usually variants of normally 10-armed 

 types. 



Since the periphery of the disk always reaches the level of the second 

 brachials of the free undivided arms, it is evident that, as the elements of the 

 division series are always of approximately the same relative size, the number 

 of arms is more or less strictly proportionate to the volume of the visceral mass, 

 and the larger the visceral mass proportionate!}', the greater the number of arms. 

 Furthermore, a very large number of arms is usually correlated with additional 

 coils in the digestive tube. But since the length of the individual arms decreases 

 proportionately to the number of arms present, due to a reduction both in the 

 number and in the length of the component brachials, the food-collecting surface 

 in 10-armed and in multibrachiate types of the same actual size can not be greatly 

 different. 



An obvious advantage of the multibrachiate condition for such species as 

 inhabit the tropical littoral is that in case of accident from storms or other causes 

 resulting in the loss of arms much less damage proportionately would be done than 

 in the case of 10-armed forms, for in one of the latter the loss of a single arm 

 would decrease the food-collecting ability 10 per cent, while with a species with 150 

 arms 15 arms would have to be lost to cause a comparable amount of injury; and 

 since the arms of the 10-armed types are relatively much longer than those of the 

 multibrachiate forms, the lesser radius of the latter enables them to escape many 

 blows which would injure the former. 



It is probably that the multibrachiate condition owes its origin or its per- 

 petuation to some factor in the ecology of the tropical littoral comatulids which 

 does not in like degree affect those living in deep or cold waters, and this factor 

 very possibly has to do it with an excess of vegetable plankton and silt in the food 

 which would require for its assimilation an enlarged digestive system, necessitating 

 an enlarged visceral mass. 



The solubility of calciimi carbonate (CaCOj), of which the skeletons of the 

 crinoids are chiefly composed, increases with the decrease in the temperature of 



