149 
of Edinburgh, Session 1875-76. 
I may say it is a year or two since I first tried the plan. 
It consists in first filling the outer glass cell one-third or one- 
half with fine silver sand, then pushing a ring of glass (which I 
cut from a common ale pint-bottle with a hot soldering bolt) 
down an inch or so into the sand. The zinc element forms a ring 
round the glass, and the copper lies as a plate on the sand within 
the glass. Its superiority to the gravity batteries, and to those, 
such as Sir William Thomson’s, where the sand forms a dividing 
layer between the copper with its sulphate below and the zinc with 
its liquid above, will be obvious. I did not find the cupric sulphate 
solution to diffuse into the zinc division. It so readily admits of 
inspection that it would be infinitely preferable' to the Meidinger 
and other plans. 
2. On the Properties of the Perigon Yersor. By G. J. P. 
Grieve, Esq. Communicated by Professor Kelland. 
3. Descriptions of some new or imperfectly understood 
Forms of Palseozoic Corals. By H. Alley ne Nicholson, 
M.D., D.Sc., F.R.S.E., Professor of Natural History in the 
University of St Andrews, and James Thomson, F.G.S. 
(. A bstract). 
In this communication the authors gave descriptions of several 
new or imperfectly understood forms of Paheozoic corals. After 
giving a general account of the method of investigation employed 
by them, the genus Heliophyllum , Hall, was discussed at length. 
The external structure of this genus is very peculiar, and it was 
shown that the genus is not by any means as nearly related to 
Cyathophyllum as has been generally believed. The new genus 
Grepidophyllum was proposed for a group of forms possessing the 
extraordinary and characteristic endothecal dissepiments of Eelio- 
phyllum, but with the remarkable character that the central tabu- 
late area of the corallum is shut in by a well-developed accessory 
wall or inner mural investment. Sometimes this secondary 
investment constitutes a complete circular sheath to the central 
tabulate area, and in this case all the primary septa become 
