(ancient) ROMAN BATHS IN BATH. 139 



molluscs of its own species. In Mr. Rimmer's book its prolific 

 character is proved by the fact that a single creature has been 

 known to deposit 1,300 eggs in one season. There is one other 

 shell in the collection which Mr. Rimmer has also named for me, 

 and this is a single specimen of land-shell, Pupa umbilicata. It 

 belongs to the third family of terrestrial shells, termed Jleliddce, 

 and the sixth genus of that family. It inhabits the crevices of 

 walls, or lives under stones and fallen leaves. This species is ovo- 

 viviparous, and the young, which seldom exceed 5 in number, 

 often remain for a time attached to the shell of the parent. The 

 shell of immature specimens of this mollusc is frequently so 

 unlike that of the adult, that it may be mistaken for a distinct 

 species. The only remaining shells that I have discovered, are 

 those which I have already referred to as belonging to brackish 

 water, Hydrobia ventrosa. They are very pretty, and were found 

 on the outside of the bath, and not in the mud bank, but I found 

 them in very considerable quantities. The mud itself, after 

 boiling in acids, has yielded a fair proportion of Diatom valves : — • 

 Cocconeis placcntula, Naviaila, Piimularia viridis, and Cydotella^ 

 etc. : all of which are fresh-water species, but I have also found 

 fragments of Coscmodiscus, which are only marine, and on one of 

 the slides prepared from this deposit there is a perfect circular 

 valve which can only be compared with Coscinodiscus minor^ 

 which is, however, a fresh-water species. The question again 

 arises. How came the marine species in the deposit? 



Passing to the stratum immediately above the firm mud, which 

 was purely vegetable, I have no hesitation in stating that the 

 curiously serrated forms upon the slides which I have prepared are 

 the siliceous cuticles of the Dutch rush {Equisetum hyemale), and 

 probably those belonging to the Avena fatua^ or wild oat, accord- 

 ing to type slides which I have procured for comparison. The 

 deposit has been treated with acid. In many of the slides there 

 are numerous vegetable spores, which I am not able to identify as 

 spores from the Equisetum. In the upper portion of the bank, 

 the light greenish tinged deposit appeared, under the microscope, 

 as before stated, to be composed of cylindrical and hollow organ- 

 isms. After treating this with nitric acid in a boiling condition, 

 the residue was found to contain a considerable quantity of dia* 



