WITH NOTES ON THE WEST INDIAN SPECIES. 285 



TORNARIA. 



A Tornaria which I obtained in the tow-net in Blanche Bay and off the small 

 coral island (Pigeon Island) referred to above, belongs to the group of Tornariae in 

 which the longitudinal ciliated bands are drawn out into tentacular processes. 



Spengel calls all such tentaculated Tornariae T. grenacheri, and speaks of their 

 circumtropical distribution (circumterrane Verbreitung) since they have been found in 

 the tropical regions of all the great oceans, Atlantic, Indian and Pacific. 



Spengel does not speak in terms of absolute certainty as to the identity of all 

 the forms designated by the common name T. grenacheri, but gives it as his im- 

 pn —ion that they are so. In this impression I think he is certainly in error. The 

 name T. grenacheri obviously implies that the forms included under that name are 

 the larvae of one species of Enteropneusta. As a matter of fact there is reason for 

 supposing that Tornaria docs not voluntarily migrate far from the habitat of its 

 parent, those which are found at great distances from home having been carried away 

 by currents and doomed to destruction 1 . An instructive example of this is furnished 

 by Agassiz' Tornaria tT. Spengel) which is sometimes taken off the coast of 



Massachusetts, and was thought to belong to the Balanoglossus (B. kowalevskii) which 

 occurs on the same coast, until Morgan 2 showed that the latter had a direct develop- 

 ment and was in fact identical with the species whose development had been worked 

 out by Bateson. 



T. agassizii very possibly belongs to Pt. aurantiaca and is liable to be carried 

 up north by the Gulf Stream. 



Pt. biminiensis n. sp. (see below), whose development has been studied by Morgan, 

 posscssi- a Tornaria ot the tentaculated type. 



From the small size of the egg it is, I think, quite certain that all Ptychoderidae 

 develop indirectly with a Tornaria larva, and it is probable that the Spengelidae do 

 the same 3 . On the contrary there is no room for legitimate doubt, in consideration of 

 the size of their eggs, that all Balanoglossidae develop directly without a Tornaria. 



The Tornaria from New Britain which is here figured most probably belongs 

 either to Pt. carnosa or to Pt. ruficollis. but it is impossible to say which, because, 

 although I obtained at least one specimen immediately after the metamorphosis, it is 

 notoriously impossible to identify newly metamorphosed larvae of Enteropneusta. 



No doubt the differences between the Tornariae of some species are very trifling, 

 but it is a great mistake to imagine that all tentaculated Tornariae belong to one 

 species. 



1 This has also been found to be the rule for Echinoderrn larvae (Th. Mortensen, Die Echinodermenlarven 

 der Plankton-Expedition, 1898). 



- T. H. Morgan, "Balanoglossus and Tornaria of New England," Zool. Am. xv., 1892, p. 456. 

 3 The diameter of the ripe eggs of Glandiceps hacksi is about -1 mm. (Spengel). 



