300 E. A. MINCHIN. 



Since these differences in two such closely allied forms must 

 be the result of different adaptation to circumstances, we 

 ought to be able to find in the life conditions of these forms 

 some explanation of it, and I believe that such an explanation 

 can be given. As I have shown elsewhere/ Holothuria 

 nigra possesses immense numbers of Cuvierian organs which 

 are thrown out on stimulation, and in the ejection the wall of 

 the rectum is ruptured, so that the contents of the body-cavity 

 can escape. If now G. irregularis lived in stalked vesicles 

 while still simply in the adult form, it would often run the 

 risk of having the stalk of the vesicle ruptured, and being 

 thrown out before its time. When once it is encysted, how- 

 ever, it may be of advantage to be thrown out, and thus escape 

 to the exterior. On the other hand, Holothuria tubulosa 

 is entirely without Cuvierian organs, hence G. holothurise 

 does not run the same risk of being cast out. In fact, it must 

 be rather hard for the spores in this form to escape, and I am 

 rather inclined to attribute to the caudal process some action 

 in the escape of the spores from the body-cavity, perhaps 

 being used to bore through the tissues in some way. On the 

 other hand. Semper (11) has suggested that the autotomy of 

 Holothurians may be of biological importance to the parasitic 

 Entoconcha mirabilis in their body-cavity, the young of 

 which could become free by producing autotomy through 

 internal stimulation, and the same might apply to the Grega- 

 rines. The irregular form is perhaps due to life in the com- 

 plicated anastomosis of the blood-plexus, and is only retained 

 as long as the animal continues to live in the lumen of the 

 blood-vessel, i. e. throughout life in G. irregularis, but only 

 in the young stage of G, holothurise. Thus the difference 

 in form would also be indirectly caused by the presence or 

 absence of Cuvierian organs. 



It is obvious that if a genus is to be constructed for these 

 forms, it must not be defined by characters that are not con- 

 stant in all the species. Armed with this axiom, let us con- 

 sider previous attempts to classify these organisms. 



' ' Annals and Magazine of Natural History,' October, 1892. 



