ACTINIAN BUNODEOPSIS GLOBULIFERA. 309 



yet immature, but no larger were to be met with, and, judging from the number 

 of tentacles and the dimensions, Verrill's Bermudas examples were in much the same 

 condition as the Jamaican. 



Undoubtedly the imperfect development of the tentacles within a restricted area, 

 noticed among the external characters, is determined hy the imperfect growth of the 

 mesenteries. The two are associated phenomena; the rate of tentacular growth is 

 regulated by that of the mesenteries, and where the latter are irregular the tentacles 

 will be the same. 



This regional irregularity of groAvth of the polyp would appear to be connected in 

 some Avay with an asexual method of reproduction. In the more common B. antilUensis 

 I have found that new polyps are produced by the process which Andres (1881) has 

 described as " Scissiparity." Fragments of varying size become gradually detached from 

 around the base of the column, attain an independent existence, and develop into new 

 polyps with tentacles and oral disc. Andres has followed the process of development of 

 the new^ individuals as it occurs in Aiptasia lacerata, and in this species the normal 

 hexameral disposition of the organs appears to be early assumed. I have not observed 

 scissii^arity in B. globulifera, but the other Jamaican species presents exactly the same 

 tentacular and mesenterial irregularities. 



Irregularities would be much more likely to occur in polyjjs thus originating than in 

 individuals developed directly from the larva. The two species of Bunodeopsis occur 

 in large numbers wdthin restricted localities, and such a distribution is mo.st probably 

 the result of a like method of reproduction in each. The West Indian Aiptasia tagetes 

 (Duch. & Mich.) also reproduces in a most in-olific manner by scissiparity, and an 

 examination of its mesenteries and tentacles reveals many departures from the hexameral 

 regularity. McMurrich (1889), on the other hand, describes the Bahamas representa- 

 tives of this species as hcxamerous, without mentioning any asexual reproduction. 

 G. H. Parker (1899), who has made a detailed study of the effects of longitudinal fission 

 in Iletridium marf/iiiafum, finds the mesenteries are very irregularly arranged in what 

 may be assumed to be fission polyps. 



Returning to the individual mesenteries, the members of the first order are found 

 to extend the whole length of the stomodseum, and some are continued as far as the 

 base of the polyp, across which they extend to the middle (PI. 25. fig. 4i). The complete 

 members of the second order may become free before reaching the stomoda^al termination. 

 At first the free edge of the mesenteries is straight, but soon it becomes greatly folded 

 and contorted, some mesenteries beconiing much more complex than others, so that they 

 may appear several times in the same section (fig. 4). The incomplete mesenteries 

 first appear in the uppermost part of the column-w^all, but all except the youngest 

 are better developed in the middle part of the polyp. Very often the mesenterial 

 mesogloea and that of the column-wall are seen to be interrupted, the continuity of the 

 circiJar endodermal muscle of the colvimn-wall being thereby preserved (PI. 26, fio-. 17). 



"Within the stomodoeal region of most polyps the mesenteries are extremely narrow, 

 and, except at their origin in the column- wall, the musculature is so weak as to 



