20 Messrs. Richardson, Nicol, and Parnell on the 
It will be seen that the agreement of the numbers in 
both columns III. and V. is extremely good. The extreme 
difference in column III. is only 5 per cent., whilst the agree- 
ment in column V. is still better, the greatest difference 
amounting to about 2 per cent. The mean value of A from 
this series of observations was found to be 3°924 x 10—* gms. 
per sq. cm. 
The next series of observations (Table IV.) was for a 
slightly higher temperature (1176°C.) and a much smaller 
range of pressure (13 mms. of mercury as compared with. 
32 mms. before). The agreement here is still satisfactory 
though not so good as before. his is because the limited 
range of the observations increases the relative errors. 
TABLE lV. 
ES Eerie 7 alent ake és 
1. II. eee is ie 
yea, | 9y?4+-112y?. Vic a 
46 204-7 ie 0 204-7 
31 1640 Me 37-0 201-0 
21 | 1380 t3 67D 2055 
15 | 116-4 3:44 890 | 205-4 
‘9 90:2 3:27 115-2 | 2054 
oe, 514 3°67 14911 | 2005 
| Jia 
These numbers give the mean value of A at 1176°C, as 
4°58 x 10-8. 
The next table (Table V.) gives the values at 1021°C. This 
shows the best agreement of the whole series, for the extreme 
difference in A is only about 2 per cent., whilst that of B is 
about 4 per cent. The difficulty of reg gulating the tempe- 
rature made errors of this amount unavoidable. The numbers 
give the mean value of A at 1021°C. as 1:77 x 10-8 gms. 
per sq. cm. per sec. 
TABLE V. 
Pee Il. pie ery | ae 
| ysemay, | 95y*—-112y2. 1336 ¢. B. 
106 ly Ae eee 0 | 3132 | 
91 ES) ae a 2445 | 3141 | 
7-6 264-2 wet 59°35 | 3236 | 
61 Dy eee an ee 8015 | 3164 | 
46 BOAT: Oh M92 | 3800 a 
31 1640 | 1:352 | 147°5 S116. 
21 138-0 | 1822) 7 | 3157 
15 1164 | 1330 | 207-7 3241 
ae 902 | 1334 | 2965 3167 
3 | 51-4 | 1340 | 2723 3237 
