NOTES ON VARIATION IN THE TUNICATA. 361 



marginal teeth are slight, but in others they are long tentacle- 

 like projections comparable with languets.* 



The tentacles at the base of the branchial siphon are of 

 considerable value in characterizing species, genera, and 

 families. In most Ascidians they are simple elongated pro- 

 cesses, but in the Molgulidae, and in two sub-families of the 

 CynthiidaB, they are compound and branched. In many 

 species the tentacles are of two or more sizes, and the differ- 

 ent orders are arranged with regularity. The smaller and 

 more numerous tentacles are always the most liable to 

 variations, such as suppression, reduplication, and irregu- 

 larity in position. Sometimes, in place of all the tentacles 

 springing from the same line, one order is inserted further 

 forward or further back than the others. This is sometimes 

 the case in Ascidia plebeia, f and I have recently found 

 the same condition in a Compound AscidianI (Botryllus 

 smaragdus). 



The dorsal tubercle, which is the more or less com- 

 plicated aperture of the duct from the subneural gland, is 

 a very variable organ, and must be used with great caution in 

 characterizing species. I have already discussed elsewhere § 

 the range of variation of the dorsal tubercle in some of the 

 commoner species of British Ascidians, and, in examining the 

 collection of Tunicata from Liverpool Bay, I have met with 

 some marked cases of variation in this organ. In a specimen 

 of Polycarpa pouiaria, a species in which the dorsal tubercle 

 is usually cordate in outline, it was found to be a complete 

 ring — a condition sometimes seen as a variation in Sti/ela 

 grossularia. In a specimen of Molgula occulta, again, one of 



* See PI. VI. fig. 5, I, illustrating the Eeport upon the Tunicata of the 

 L. M. B. C. district, in this volume. 



t See Journ. Linn. Soc, Zool., vol. xv, pi. xix, fig. 4. 



\ See PI. VI., fig. 7, illustrating the Report upon the Tunicata of the 

 L. M. B. C. district, in this volume. 



§ Especially in Proc. R. Fhys. Soc. Edin., vol. vi, p. 250. 1881. 



