Chemistry and Physics. 217 
and breadth are 55 cms., and its height 60 cms. The a is 
however much less, being only 290 Carcel — instead of 
1,000. A greater light may nevertheless be obtained by an in- 
creased speed, as is shown by the following ae of. ten series of 
experiments: 650 turns gave a light of 77 puannng 850 gave 125; 
880, 150; and 900, 200 without heating or spar 935 gav e 250 
with slight heating, and 1,025 gave 290, but produced heating 
and sparks. Comparing with the machine of the co pany of the 
pies a space of 170135150 ¢ Its a is therefore 
twelve times as great, the siirtace | cote a seventeen times as 
great, and the volume eighteen times as great. 
These machines having neither cranks, connecting shafts or 
dead-points, are well adapted to the transformation of electricity 
into work. One of the small machines gave the results shown in 
the accompanying table, when run with Bunsen cells of 0:20. 
The first column gives the number of cells, the second the number 
Cell. Turns. Power. Cell. Turns. Power. 
2 760 (: 1100 4- 
$ 810 1°02 8 1100 5°00 
4 1000 1:02 8 900 4°81 
4 900 1°80 9 1500 Tog a} 
5 1100 2°50 10 1700 5°52 
6 1000 3°36 10 1300 6°16 
kgmtrs., determined by a friction brake. The current me a sec- 
than ha ower was first t changed tate giesuicite 
and then the electricity into sade: the sphere of efficiency of 
eac = t have been over 70 per 
one idee the other fine. Ba current from two Bunsen cells is 
then passed rg Ceagns the first, running the machine, and producing 
a current of high tension in ie fine coil. A telegraph may thus 
é run ies two Bunsen cells.— Compt. Rend., |xxix, 1178. —— 
fully illustrated on the same subject may also be found in Engi 
— Nov. 27th, 1874, and in the Annales Industrielles, B. ©. P. 
nsion of Phos sphorus.—Mus. Prisati and DrFRancuis 
are measured the expansion of both solid and liquid phosphorus 
at temperatures from 0° to 270° C. They employed a cylindrical 
dilatometer with a graduated neck. This was imme a ba 
of oil kept stirred continuously, and the greatest errors were esti- 
mated at atenth of a degree. The readings were all reduced to 
egrees of the air-thermometer by a direct comparison, From a 
