e 
20 B. Silliman on the relation between the intensity of 
2d¢ Trial—Consumption of sperm, 32°2 grains. 
“ ‘gas, 4°58 cubic feet. 
Mean candle power of 15 observations, 11°8 candles. 
The above data caleulated by Farmer's Theorem, 
5-004 cubic feet, and 32-7 grains give 15°15 candles. 
4°58 6c oe 32° “é 66 15°09 “6 
Difference, a. eg 
Calculated by the old rule. 
5°004 cubic feet and 32°7 grains give 15°16 candles 
4°58 6 os 22°93 Se 46 ac 18°82 ‘ 
Difference 1°34 sy 
It is obvious from the study of these results, that within the — 
limits named the increase of intensity in gas flames, whether 
naked or argand, is at a ratio certainly as great as the squares _ 
of the volumes of gas consumed ; and hence it follows that 
e 
all the photometric determinations, which have been obtained © 
by computation from volumes greater or less than the assumed _ 
standard of five cubic feet per hour, in the simple ratio of the — 
volumes consumed must be considered as absc lutely worthless, 
provided the theorem of Farmer here announced is confirmed. 
It is evident also that this theorem applies with equal force _ 
io the weight of sperm consumed by the standard candle as to 
_ the volumes of the gas burned in equal times. 
following trials exhibit the result obtained by burni 
re om saeabesaines 
Heh RE Oa ie 
urning smaller 
values obtained by the two 
patie 
