333 
of Edinburgh, Sessio?i 1876-77. 
(L), taken in air at different pressures (P), between two metal balls 
of 7 mm, 5 radius, the requisite potential (V), is expressed by the 
formula 
V ccP JL. 
The Holtz machine employed is a double one, made by Ruhm- 
korff, and it was used with its small Leyden jars attached. The 
measurements had to be made with a divided-ring electrometer, 
so that two insulated balls, at a considerable distance from one 
another, were connected, one with the machine, the other with the 
electrometer. With P constant the curve for V closely resembles 
a parabola between L = 0 and L = 60 ram , but for higher values of L 
it appears to tend towards an asymptote at a finite distance from the 
axis. Thus it would seem that to double the length of very long 
sparks under these circumstances, a comparatively small percentage 
increase of potential will suffice. This may enable us to explain 
some singular peculiarities of lightning. Mr Macfarlane intends 
to work up this subject very thoroughly, with the help of Thom- 
son’s Long Range Electrometer. 
(6.) The Thermal Conductivity of Gas Coke. By Messrs Knott 
and Macfarlane. 
The method employed was the same as that described (Proc. 
VIII. 623) , but the high conducting power of gas coke for elec- 
tricity made the experimental work very difficult. The results, so 
far, are not very consistent with one another, but they appear to 
point to a diminution of conductivity by rise of temperature. 
(c.) Preliminary Experiments on the Currents produced by contact 
of Wires of the same metal at different Temperatures. By 
Messrs Knott and Macfarlane. 
These experiments, so far as they go, confirm the results 
formerly obtained by Mr Durham, Proc. VII. 788. 
(d.) On the Relative Percentages of the Atmosphere and of the 
Ocean which would flow into a given Rent in the Earth’s 
Surface. By Professor Tait. 
The note had reference to some sensational statements recently 
