W. G. Van Name — Bermuda Ascidians. 363 



striking variations as subspecies. Between these there are an indefi- 

 nite number of intermediate forms. Nevertheless all these forms 

 appear to have some degree of permanence, and reproduce their 

 peculiarities at least in their immediate descendants, for often a 

 number of colonies attached to the same stone, or growing near 

 together, will have exactly identical characters, indicating a common 

 parentage, while certain others, growing among them, will differ 

 from them, yet agree among themselves. 



The relation of these numerous varieties to previously described 

 species of the genus is a difficult question. Naturally their nearest 

 allies would be sought for on the Atlantic coast of the United States 

 and in the West Indies. In neither of these regions has the genus 

 been sufficiently studied. 



Sluiter (15) has recently described two new forms, L. conchylhi- 

 tum and X. cineraceum, from Jamacia, but these differ from the 

 Bermuda forms, among other things, in the number of stigmata, for 

 he says that they are provided with but from four to six stigmata in 

 a row on each side, while the Bermuda forms have about 1 2 in those 

 with lai-ge zooids, and probably at least 8 or 10 in all cases. 



From the Atlantic coast of North America two species only have 

 been described, as far as the writer is aware ; L. albidum Verrill 

 and L. luteolum Verrill, the latter perhaps only a variety of the 

 former. Both of these are found on the New England coast. The 

 Bermuda varieties are quite different from the typical albidum, 

 which has spicules of a different type from any of the Bermuda 

 forms (fig. 41), and in most specimens the spicules are much larger 

 than is the case in any of the latter. 



Specimens of L. luteolum, from Southern New England, however, 

 have spicules more like some of the Bermuda varieties (fig. 40). 

 Yet the correspondence between L. speciosum Herdman (6), from 

 Bahia, Brazil, and the commonest Bermuda form is so much closer 

 that it seems best to consider the latter, and consequently the 

 remaining Bermuda varieties, as subspecies of the Brazilian form. 



The writer has not ascertained that any of the several varieties 

 here described is confined to any particular locality at Bermuda. 



Leptoclinum speciosum Herdman. 



Leptoclinum speciosum Herdman, Report Voy. Challenger, pt. xxxviii, p. 274. 



The types are from Bahia, Brazil, in 7 to 20 fathoms. I have not 

 found specimens at Bermuda which correspond exactly to the 

 description of the Brazilian examples. 



