BOYHOOD 19 



coining when in a humour to use him. Should this time, how- 

 ever, be too short for thy satisfaction, tell me, and I shall " sup- 

 press my feelings." The reason I want him back is, that I amuse 

 myself of an evening with writing my " Bivalvia Hibernica," 

 being a description of our native shells. I have got as far as 

 Donax, so am nearly half through. 



To the Same. 



January, 1831. 



I have thy letter, with Fleming, and the lichen (aquaticus ) 

 which I hope to examine with the microscope to-morrow. I 

 ha7e little doubt but that it will prove to be JRiccia natans (one 

 of the Hepaticse, and first cousin to Jungermannia), rather a rare 

 plant, never before seen by me, and whose fruit has not yet 

 been detected in these islands. I shall rejoice to receive more 

 specimens, enclosed in a small phial of water. When sending 

 the specimens, state the exact habitat, whether common, and other 

 particulars. 11. natans is described by Hooker and Taylor to 

 possess long flat latinise, mostly purple at the margin and 

 underneath. I find such on thy specimens, but they are green, 

 not purple. Try if any purple can be found. If the plant be 

 really R. natans, it may open a communication between us and 

 Hooker. I mean to send the Potamogeton oblongus 1 to 

 Bicheno, in the spring. I am pretty sure it is the oblongus, but 

 nevertheless only a variety of natans occasioned by shallow 

 water, which might render the leaves less coriaceous and more 

 membranaceous. 



I do not feel competent to speak on the genera or families 

 of univalve shells, as I know so few of their animals ; and from 

 those I do know I am convinced that no system in accordance 

 with nature can be established, where the generic characters are 

 taken exclusively from the shells. In the case in question, viz., 

 univalve shells, I call no genus natural except that in which the 

 animals have the same external and internal characters. 

 Wherefore I hold, by easy extension of this principle, that the 

 most natural families will be founded on the animals. Cuvier 

 classes the univalve mollusca by the difference in their breathing 

 apparatus, to me, very properly. Turbo elegantissimus affords 



1 Pond weed. 



c 2 



