338 H. Wurtz on a Gas Well in New York. 
was not constant, showing a heating of eb iron by radiation and 
conduction from the flame above. At one time, however, the 
temperature sunk to 59° F., so that the newt temperature is be- 
low this, it may be as low as 50°; very low for a depth of 500 
feet. Doubtless the gas is cooled by expansion from a state of 
compression. (A curious suggestion occurs here regarding cau- 
ses of irregularity in increase of temperature in descent in some 
fossiliferous rocks. 
The candle power was sed oe with a standard candle by 
contriving a small dark room with a large blanket shawl, usin 
of course the hr or i shdoW test. The gas was burn 
from a es steatite-tip bats-wing burner, being first passed 
through a glass tube so stuffed with cotton as to reduce the 
pressure mit 6 that which gave the maximum of light. The 
result was about sz candles. I had not with me an Argan 
burner, which, renee e ~ = a very contracted throat, would 
doubtless afford wi a éonsiderably higher candle 
power. It is well I known t i the effect of carbonic acid in 
iminishing cee power is very far less in the Argan 
thats in flat flame burn 
e con tion lost wid made by immersing in snow and 
salt, in a common water bucket, some sixty feet of small india 
rubber ee that I had with me. The thermometer stood at 
o change of the candle power pa 
during half an ho atid wens the light-giving hydrocarbons 
+ gases, or at least a coally i in- 
condensable. Lime-water shave carbonic acid to be largely 
ese 
Manhattan Gas Light Co., in v this i ty, and some a analyses made; 
3 manipulations being used which were devised 
by myself, in SSajrinetion with Professor Silliman, and which we 
have not yet publis 
Results of the analyses. 
_ Marsh gas a ee 
SE TY Dear ere tie acne aes Cee goes 10°11 
Nitrogen : omit 
_ Oxygen -.. 2 eS 
Tluminatin i oe 
: 100-00 
