354 
DR. MEYER WTLDERMAN OX CHEMICAL DYNAMICS 
a metal plate on the india-rubber ring somewhat loosened. Water entered the 
container, the tap was turned off, and the acetylene with the admixed air allowed to 
escape between the metal plate and the india-rubber ring. This was repeated two or 
three times. In this way the air was expelled from tire carbide containers so that 
when the cross-bar was tightened up and the tap opened again, no appreciable effect 
of the new container upon the acetylene ought to have taken place. With such 
precautions this source of error quickly disappears, owing to the gas-holder, leading 
pipes, balance governor, purifier, and burner containing a great quantity of pure 
acetylene—as can be seen from the observed intensity of light read from the deflection 
of the galvanometer. Variations, however, still existed, and sometimes required to 
be adjusted by the regulating tap. Another possible source of error was the variation 
in the composition of the gas, arising from the carbide used not being always of the 
same quality (it is not certain that such a source of error does exist). This was 
counteracted by the adjustment of the regulating tap. 
These adjustments were always guided by the indications of the deflection of the 
galvanometer. 
Water Manometer. (See fig. 5.) 
Since it was necessary to work with a pressure of about 4 inches when the tap of 
the burner was shut, and of about 2\ inches when it was open, OT millim. variation 
in the height of the upper and lower side of the water manometer would indicate 
0'2 millim. in 25 millim., i.e., 0‘8 per cent, variation in the pressure of the gas. Its 
chief purpose is to indicate in a quick manner, whether all the apparatus connected 
Avith the supply of the acetylene gas to the burner is in good working order. 
The Acetylene Burner. (See fig. 5.) 
All attempts to get a light of a great candle power, which would remain of a 
constant intensity and composition have been, as far as I know, unsuccessful up to 
the present. We have now standard lights of 1 or 10 candle power, but we have 
none of 200 or 1000 candle power, since the light of the arc varies considerably 
both in intensity and composition. The object here was to obtain a light of, say, 200 
or 500 candle power, or of any other intensity desired, which would remain constant 
in its intensity and composition for any length of time, which could at any time be 
easily adjusted with great accuracy to the desired intensity, and which could be used 
with ease in the ordinary work of a laboratory. With the burners on the market 
which were tried, the gas is always passed either through a very thin slit giving a 
flat flame, or through two thin pinholes—the two gas streams meeting in one point, 
and giving again a thin flat flame (24-30 candle power). This is done in order to 
get the surface of the flame in contact with the air as large as possible, to obtain 
complete combustion and as white a flame as possible. It was found that after 
