206 Prof. J. Dewar and Mr. A. Scott .
[ June 19 , XVII .
" On the Vapour Densities of Potassium and Sodium .
" By Professor James Dewar , M.A. , F.R.S. , and Alexander Scott , B.A. Received June 12 , 1879 .
The ingenious method of determining the vapour densities of volatile bodies invented by Professor Meyer* suggested the revision of work done by one of us in connexion with Professor Dittmar , f on the vapour density of potassium .
The new method possesses several advantages over that formerly employed , and seemed to be especially fitted to simplify the difficulties of working with substances having the chemical characters of the alkali metals .
The fact that a constant temperature need not be maintained unless for a short time , and that all absolute determinations of temperature are unnecessary during the course of the experiments , renders Professor Meyer 's system the simplest when the boiling point of the substance is above a white heat .
The apparatus used was simply a reproduction in wrought iron of .
Professor Meyer 's glass one , on a somewhat larger scale .
The bottle A had a capacity of 160\#151 ; 180 cub. centims. The tube B was about 2 feet long and in diameter about *5 inch .
A and a considerable part of B was coated with a glaze of borax and then fitted between two crucibles as in figure , and the crucibles filled with fine sand as completely as possible .
This apparatus kept a tolerably constant temperature for a considerable time even when the furnace temperature varied considerably ) as it was only slowly affected , owing to the mass of the crucibles and sand surrounding it .
The chief difficulty at first was the diffusion of the furnace gases , especially carbonic oxide , through the hot iron .
This , however , seems to be completely prevented by the borax glaze .
To ensure success , especially in the cases of potassium and sodium , it was found to be essential to clean the apparatus by treatment with dilute hydrochloric acid to remove the larger part of the oxide of iron , then to fuse in it some caustic potash , and thus get rid of the silica , \amp ; c. , used in the welding in of the tube and bottom of A , after boiling several times with water to dissolve out the potash , it was again treated with dilute hydrochloric acid , and then thoroughly washed and dried .
The crucibles were arranged of such a height that the iron was in no place exposed to the gases of the furnace .
The fuel used was gas carbon broken into pieces the size of walnuts , this gave a very constant temperature , and the furnace fed with it on more than one occasion fused the bottle A. * " Deut .
Chem. Ges .
Ber .
, " vol. xi , 2253 .
f " Proc. Roy .
Soc. , " vol. xxi .
1879 .
] On the Vapour Densities of Potassium and Sodium .
207 After the apparatus had been placed in the furnace and had risen to a dull red heat , the tube E was inserted into it and made airtight at C by the india-rubber tubing F , the point of the tube reaching almost to the bottom of A. Hydrogen was now passed into the apparatus by D , and a current kept up until the apparatus was thoroughly reduced ; by this time the temperature had risen sufficiently for the experiments to be proceeded with .
The apparatus *\gt ; was now filled with nitrogen , which was likewise passed in at D , the tube E withdrawn , and an india-rubber stopper inserted into C , and the end of I ) immersed in water or mercury .
The weighed substance was then dropped in at C , and the gas expelled , collected in a graduated jar in the usual manner .
The substances whose vapour densities are to be determined , were placed , after weighing , in small iron capsules and kept in them by a plug of wire gauze .
This was done in order to ensure their reaching the bottom of the apparatus , and the wire gauze plug also had the advantage of moderating the rate at which the gas came off .
The potassium and sodium were placed in tared iron capsules and weighed in dry carbonic acid , and then the plug of wire gauze inserted .
To test the apparatus cadmium was tried and gave very good 208 On the Vapour Densities of Potassium and Sodium .
[ June 19 , results .
Zinc was next used , but with very slight success , owing , we believe , to the formation of an alloy of zinc and iron , even although the temperature was far above the boiling point of zinc .
This was shown by the wire gauze plug and capsule becoming very brittle , so that the wire gauze might be rubbed to pieces between the fingers .
An approach to approximate results was once got with an apparatus which had not been reduced by hydrogen , the coating of oxide of iron and silica having protected the iron from the zinc .
Mercury gave good results in this apparatus .
With potassium and sodium greater difficulties stood in the way of exact results , as the correct weight of the pieces inserted could not be ascertained , and both metals seemed to have a tendency to behave in the same manner as zinc , viz. , to form an alloy with the iron ; because the vapour evolved after a few minutes was absorbed completely in spite of all precautions to prevent oxidation .
Both of these sources of error tend to make the numbers found for potassium and sodium too high .
The mean error due to the weighing of the metals was found by measuring the hydrogen given off by their action on water , and the action of the iron was partially overcome by raising the temperature as much as possible , so as to induce rapid volatilization .
After each experiment the metallic vapours were blown out by a strong current of hydrogen , by inserting ( as before in reducing the apparatus ) the tube E , and passing the hydrogen in at D , refilling with nitrogen , and proceeding as before .
The following molecular weights are calculated by finding the weight of metal in milligrams required to give 22"34 cub. centims. of vapour at 0 ' 0.+ 760 millims. pressure .
Cadmium .
OTOO grm. .
gave 22*0 c.c. at 0 ' + 760 mm. . . .
mol .
wt. = 107*7 0-103 99 19-4 99 99 " 109-6 0-106 99 20-8 99 99 " 113-2 o-ioo 99 19-2 99 99 116-3 0-105 99 20-4 99 99 " 115-0 0-102 99 20-7 9 9 99 " no-1 0-102 99 20-9 99 99 1090 0-106 99 211 99 99 " 114-0 894-9 Mean molecular w eight .
111-86 Mercury .
0-198 grm , .
gave 22'8'N3.c .
at 0'C .
+ 760 mm. .
- .
mol .
wt.=194 0*210 99 23-8 99 99 " 199-6 1879 .
] The Physical Properties of Liquid Acetylene .
209 Potassium .
0"065 grin , gave 18*8 c.c. at 0'C .
+ 760 mm. .
mol .
wt.= 77'2 0-083 " 22-4 " " " 82-7 0-077 " 19-6 " " " 87-7 0-075 " 18-5 " " " 90-5 0-077 " 20-0 " " " 86-0 0-070 " 20-0 " " " 78-2 0-077 " 22-6 " " " 76.1 * 0-087 " 26-1 " " " 74-4 652-8 Mean molecular weight ... ... ... 81'6 Sodium .
0 043 grm. gave 15*1 c.c. at 0'C .
+ 760 mm. .
* .
mol .
wt.= 63*6 0-067 " 23-9 " " " 62-6 0-047 " 19-2 " ' " " 54-7 0-040 " 17-6 " " " 50-7 0-059 " 28-6 " " " 46-0 \#171 ; 0-040 " 20-4 " " " 43-8 0 045 " 23-36 " " " 430 0-037 " 20-2 " " " 40-9 405-3 Mean molecular weight ... ... ... 50*66 Four pieces of potassium weighed in iron capsules and thrown into water , gave 84'2 as a mean molecular weight , calculated from the hydrogen evolved .
Four pieces of sodium similarly treated , gave 49'8 as a mean .
The above experiments show that no great accuracy in the determination of the vapour density of sodium and potassium can he attained by this method of working* when vessels of wrought iron are employed .
The results are , however , conclusive as regards the normal character of their respective vapour densities .
XVIII .
" The Physical Properties of Liquid Acetylene .
" By Gerrard Ansdell , F.C.S. , Chemical Assistant to the Royal Institution of Great Britain .
Communicated by J. Dewar , F.R.S. , Professor of Chemistry , Royal Institution .
Received June 12 , 1879 .
The hydrocarbon acetylene , being the ojriy one of its numerous class which can be formed synthetically by the direct union of its

