460° REPORT— 1883. 
The method finally adopted was to dissolve the silver permanganate in nitric 
acid by the aid of a reducing agent, and then to determine, by the method of Stas, 
the quantity of pure potassium bromide required for complete precipitation of 
the silver in solution. The reducing agents which gave the best results were 
potassium nitrite and sodium formate ; sulphurous acid was found to leave after 
reduction always a trace of what was apparently sulphide, and was besides 
troublesome from the insolubility of the silver sulphate produced. The results of 
the titrations are given below in Table II. 
TABLE II. 
AgMn0O,. KBr Equivalent 
No. | AgMnO, | Corrected KBr Corrected of Reducing Agent 
for vacuo for vacuo | AgMnO, 
1 6°528 65289 374228 3°42385 227-094 | Sulphurous acid 
2 75368 75378 39541 3°9553 226°958 | Nitrite of Potash 
3 6°1000 6°1008 3°20067 3°20166 226°937 ns 2 
4 57457 574647 300584 3 00677 227-606 | Sulphurous acid 
5 671651 6°16593 3°23503 323602 226°918 | Formate of Soda 
6 51126 5°11329 2°68216 2°6828 226-984 » 9 
iy 50737 507438 2°6614 2°66204 227:013 | Nitrite of Potash 
8 | 13-4466 | 134484 705385 705602 226983 5 3 
9} 12°5782 | 125799 659861 660065 226°972 | Hydrogen 
10 | 12°2686 | 12:27025 64361 6°43808 226976 | Nitrite of Potash 
The mean atomic weight of manganese which results from the average of the 
eight determinations in which sulphurous acid was not employed as the reducing 
agent is 55°038, oxygen being taken as 16 and silver as Stas’s value, 107-93. 
Thus another element is added to the list of those whose atomic weights have 
been found on revision to be exceedingly near whole numbers. 
6. On the Molecular Weights of the Substituted Ammonias. By Professor 
James Dewar, M.A., F.R.S., and ALEXANDER Scott, M.A., D.Sc. 
It seemed to the authors that by the use of the haloid salts of the substituted 
ammonias, small differences from whole numbers in the atomic weights of hydrogen 
and carbon would be easily revealed. The difficulty of obtaining perfectly pure 
substances for such work, together with their hygroscopic properties, introduces 
serious difficulties; and for the purpose of testing the accuracy of the proposed 
method, the preliminary experiments have been made with triethylamine. 
The triethylamine was prepared from tetrethylammonium bromide, by dry dis- 
tillation, and the base separated as hydrochlorate. The free base was obtained by 
treatment of the hydrochlorate with caustic potash, dried carefully, and distilled. 
The portion boiling between 90° and 91° C. was converted into hydrobromate, and 
its equivalent relation to silver determined, after the method of Stas, with the 
following results :— 
Weight of Salt Weight of Silver Molecular Weight 
in vacuo in vacuo of N(C,H;)3;H Br 
6°6248 3°9219 182°315 
8:24088 48798 182-270 
A portion of the same fraction was treated with nitrous acid, to eliminate primary 
and secondary monamines, and the titration repeated with the following results :— 
Weight of Salt Weight of Silver Molecular Weight 
* in vacuo in vacuo of N(C,H;);HBr 
5°3165 31519 182-052 
4:6237 274194 182-001 
The result of the nitrous acid treatment has been to lower the molecular weight, 
