410 Proceedings of Royal Society of Edinburgh. [july 18, 
5. Od some Vapour Densities at High Temperatures. By 
Alexander Scott, M.A., D.Sc. 
The apparatus used in the following determinations is that of 
Victor Meyer, modified as described in a paper published in 1879, 
“ On the Vapour Densities of Potassium and Sodium,”* by Professor 
Dewar and the author. In the experiments here described, the 
platinum vessel was further protected by a casing of iron, and the 
intervening space filled with magnesia, the iron casing being em- 
bedded in sand enclosed between two crucibles. In Series I. 
hydrogen was the gas used to fill the apparatus, hut in all the 
others nitrogen prepared from the atmosphere by mixing it with 
ammonia and drawing the mixture over red-hot copper, then through 
dilute sulphuric acid into large glass gasholders, from which it 
was expelled after drying into the vapour-density apparatus. The 
nitrogen thus prepared almost invariably contains a small quantity 
of hydrogen, and this is a decided advantage for most of the sub- 
stances used. The temperature of the furnace (which was an 
ordinary wind furnace, with 35-feet draught, and fed with coke) 
was considerably above the melting-point of cast iron, but was 
barely hot enough to volatilise rapidly potassium iodide and silver 
chloride, and this gives their results rather high. The best way of 
weighing the potassium and sodium was found to be to cut rapidly 
a piece of the required weight as nearly as possible, and instantly 
wrap it up in a tared piece of thin platinum foil and weigh again. 
The sodium kept remarkably well thus, but the potassium was not 
so satisfactory. To check the weights, pieces were similarly weighed 
and thrown into water, when it was found to require 23'8 to 24 - 3 
milligrams of sodium to give 11T6 c.c. of hydrogen, according as 
an ordinary Becker’s balance, turning with a milligram was used, 
or a finer one turning with T milligram, the more rapid though 
coarser one giving thus the best results ; of potassium similarly 
46 ‘3 milligrams were required. 
The weight in milligrams of the substance which would be required 
to give 22 '33 cubic centimetres of vapour is taken as the molecular 
weight. 
* Proc. Roy. Soc. Lond., vol. xxxii., 1879. 
