4° 



while the filaments of the fourth couple are al most straight, an indication that the) are destined 

 to be fertile or ma mie in character. Add to these indications the facts ili.it in the older 



larvae studied by van Beneden the fourth couple of protocnemes «rere longer than any of the 

 other mesenteries and that in the oldest individual studied by Vanhöffen these same mesenteries 

 dl) the longest of all and contained immature ova, and their identification as 

 ms fairly certain. 

 In [862 A. Agassiz described another free-swimming Cerianthid larva, which he referred 

 to the genus Arachnactis as . /. brachiolata, an identification which has remained unchall» 1 

 ever since. In 1890 M Intosh described a single specimen taken in the Bay of St. Andrews 

 as an Arachnactis, and later van Beneden (1891) gave a thorough description of a number 

 of individuals in different stages of development, collected by Bourne off the south-wesl coast 

 «'t* Ireland and referred to - /. albida. Van Beneden recognized .is the results of his studies 

 that these forms were different from those studied by Boveri (1889), and in his later paper 

 identified them with that described by M> Intosh and named them A. lloydii, on the 

 supposition that they would prove to be the young of C. lloydii. Previously to this, however, 

 Fowler had recognized the fact that the forms studied by van Beneden and also indiv- 



iduals from Plymouth which he had himself studied were different from . /. albida, and named 

 them A. bournei, which term has the priority, until it is made certain that the forms are 

 really the larva' of C'. lloydii. 



Both . /. brachiolata and . I. bournei present certain marked differences from A. albida. 



Thus, in early stages the body of albida is short and almost spherical, while in the other two 



ies it is more elongated and somewhat conical; the tentacles of albida are long and 



slender, their length being many times the breadth of the disc, while in the other two their 



'i is about equal to the diameter of the disc; and in albida the median marginal tentacle 



appears only when the fourth couple of tentacles are already rather long, while in the other 



two species the median marginal tentacle developes equally with the tentacles of the third 



couple. How far these differences may be correlated with differences in the arrangement of the 



nnot yet be determined. In the case of . /. albida, as has already been noted, 



individuals with gonads and acontia have been observed, and there seems a tolerable certainty 



that t ; nterial arrangement of these: individuals is of the same type as 



that of Pachycerianthus ; but no sufficiënt data are at present available for the 



'i J' termination of the mesenterial arrangement of either . /. bournei or ./. 



XL brachiolata. Th( oldest examples of the latter form that I have heen able 



lhc to obtain possessed only seven tentacles and the mesenterial arrangement 



arrangement ol . . 



[ , eD . was that indicated in the adjoining text-figure (Text-fig. XI). The second 

 couple of protocnemes is much more developed than anv of the other mesent- 

 eries, but whether this is a temporary condition, or really indicates that they 

 to form the teloi nemes is uncertain. 



ng aside the possibility thus indicated it seems that in the certain differen 1 

 ibove there is a valid basis for the separation of brachiolata and bournei from albida, 

 ■ • A to be described below, resembles albida in the particulars in question. 



4° 



