Division in the Echinodermata. 327 



groups of arms and have more than five arms*. The trans- 

 formation of the young six-armed individuals into individuals 

 with five arms evidently requires a previous division. They 

 can only lose the sixth arm by regenerating, after the last di- 

 vision, only one or two arms instead of two or three. 



It would be very incorrect to conclude from the constant 

 occurrence of heteractinism (and of division if our interpretation 

 of heteractinism is correct) in this tolerably long series of six- 

 armed Ophiothel(e y Ophiactines, and Ophiocomo3 } that the same 

 things occur in the other Ophiurids which have normally more 

 than five arms. These, however, are not numerous as far as I 

 know ; for, leaving out of consideration the young Aster ophy ton 

 with seven arms described by me (which is as puzzling to me 

 as it was thirteen years ago, and which is still known only 

 from a single specimen) and Asteromorpha Steenstrupii (in 

 which the six-armed state is probably only an accidental 

 anomaly), we have in this category only two species of Ophia- 

 cantha, namely 0. anomala, Sars, with six, and 0. vivipara, Lgm., 

 with seven to eight arms ; and in neither of these has any thing 

 been observed to indicate a division. The necessary condition 

 for spontaneous division would therefore seem to be that the 

 species (at least when young) should have normally more than 

 five (six) arms, although we must not conclude that it exists 

 from this greater number of arms : in one of the above-men- 

 tioned groups (Ophiocoma) it is evident that the spontaneous 

 division is confined to the young ; and it is not improbable 

 that this is the case also with the others ; but this does not at 

 present appear with sufficient clearness from the facts, and the 

 solution of this important question must be left to subsequent 

 researches upon living animals. Its importance consists in 

 the fact that if it is answered in the affirmative the laws of 

 reproduction in these Ophiurids would fall under that of alter- 

 nation of generations, the young individuals then representing 

 the agamic generations, and the adults, after division, the 

 sexual ones. 



Perfectly similar phenomena are manifested in certain As- 



* All the young individuals, however, or those which have not yet ac- 

 quired the physiognomy, coloration, &c. characteristic of the species, do 

 not present six arms or the heteractinism which is associated with that 

 number. I have already mentioned (Addit. ii. p. 14G) that of 12 speci- 

 mens, 8 had six arms, two or three of which were generally shorter and 

 thinner than the others. Of 21 specimens now at my disposal (1 of O. 

 Valencies and the rest O. pumilci) I find that, with one exception, all the 

 individuals below a certain size (4 millims. 1 have six more or less unequal 

 arms, and all those which measure 5 millims. or more have five arms. 

 The exception is a specimen with six arms rather larger than it should be 

 according to this rule. 



