476 Messrs. Haselfoot and Kirkby: Electrical Effects 
TasueE III. 
| é / 
p, ve p' 7 to Vi are | q 
| | 
86 83 3 2 83 3 | 038 
845 80 45 | 25 83 a eG ‘0076 
83°5 76 Ce ee Son 4 “78 ) 1G 
eae a he T50)) 1 | 83° 1: 5a ieee 
| 8 | g25 3:5 1 38 ae ‘014 
8 | 80 foe 006 | 88. irae 014 
9) } TOS 135 9 81 . 97 ‘012 
9] | ite 16 8 81 iT 0093 | 
| Go) Sas 17 Ly 81 2-42 033 | 
92 69 23 17 8l 7 Se 
91 72 19 15 240 94 | -013 
100 88°5 tea) ‘06 240 "61 L) ) (ORGS 
100 82-5 iS 16 240 88. | “Oli 
94 715 22-5 | 29 240 2:3 032 | 
94 65 29 44 240 93 014 | 
Table III. to a more sensitive one. In the former case it 
was found that a microcoulomb produced a throw of 6°5 
divisions, in the latter a throw of 36 divisions. The notation 
is as before, and a@ is the pressure of the air, so that the 
initial pressure is p+. 
Five other observations were taken, but they are not 
reproduced, as the deflexions ranged from only 5 to 83 
divisions on the galvanometer-scale, and the percentage error 
in reading may have been considerable. They are, however, 
of some interest, since they are fairly consistent, and they 
give values of g whose mean is ‘012. For comparison with 
previous values this may be taken as ‘01, as explained 
later. 
Ten other observations were taken. In three of these the 
deflexion was very small and not more than ‘04 mm. of air 
was present; they give values of y whose mean is 0027. In 
five others the amount of air was small, probabiy not more 
than a millimetre. They give a mean value of g equal to 
‘0057. The remaining two were made with 25 mm. and 
26 mm. of air and 92 mm. of the combined gases. The first 
produced a deflexion beyond the limits of the scale. This 
corresponds to a value of g of the order obtained in the first 
experiment of Table II. The other gave a value of g equal 
to *042. 
These tables exhibit very great differences in the quantity 
of electricity produced by the explosion, far beyond the 
possible limits of experimental errors, and these are most 
