150 Intelligence and Miscellaneous Articles. 
Hence 
k= Ze, + 0m). 
4m 
We therefore obtain & from the measurable magnitudes g,, m, 
Om, and o'm; 12 which, as estimating the accuracy, it may be 
pointed out that 3, and o’,, are both positive. 
If, now, V is the potential of the block of lead, G the constant of 
eravitation, z the vertical coordinate, then 
OV 
k = SS 
Cas 
The differential quotient ey may be calculated from the known 
dimensions of the parallelopipedon, and the position of the pans. 
Our cbservations give the constant of gravitation G, and therefore 
also the mean density of the Harth. 
We have already taken the preliminary steps for the experimental 
execution of this method. We think of using a mass of lead which 
has about twice the attraction of the ball used by Herr von Jolly ; 
the determination of & may therefore, other things being equal, be 
made with eight times the certainty. Moreover, from the consi- 
derably smaller distance of the upper from the lower scale-pans, 
13 to 2 metres, we are in a position to exchange the weights 
within a closed case by our automatic arrangement, by which cur- 
rents of air and differences of temperature are almost entirely 
avoided. We may therefore accept with certainty a far greater 
accuracy for our determinations.—Siizungsberichte der Akad. der 
Wissenschaften zu Berlin, December 1884. 
RESULTS FOR USE IN CALCULATIONS WITH MANOMETERS WITH 
COMPRESSED AIR. BY EH. H. AMAGAT. 
I gave some years ago the numerical results necessary for 
calculations with compressed air between 20 to 30, and 430 
atmospheres ; but these results were not reduced to the standard 
pressure as starting-point, partly because my series began at 
pressures higher than those at which Regnault’s series end, and 
partly because these latter series were obtained at a temperature 
which is too different from that at which I worked. In all the 
researches in which I have hitherto determined the pressure with 
a nitrogen-manometer, I took this difference into account, by a 
method which it is needless to describe here. On the other hand, 
the results obtained by M. Cailletet and by myself exhibit such 
differences as to make a verification necessary. 
