CHBT:K EXPERIMENTS. 101 



an experiment is placed in a glass bottle fitted with a two-hole rubber 

 stopper. Bent glass tubes passing through the holes in the stopper 

 provide the means for forcing out the alcohol by blowing air into the 

 bottle, as in an ordinary laboratory wash-bottle. The tube through 

 which the air enters the bottle is connected with a chloride of calcium 

 tube to remove moisture from the air blown into the bottle. The exit 

 tube from the alcohol bottle is bent downward and drawn out to a 

 point and so placed that it delivers the alcohol into a burette attached 

 to the outside of the rear wall of the chamber. The burette is con- 

 nected at the bottom by means of rubber tubing, and a glass tube 

 through a cork in the outer door of the food aperture, with the long 

 rubber tube leading to the alcohol lamp. A screw pinchcock controls 

 the flow of alcohol out of the burette. 



At the beginning of an experiment the observer, by looking through 

 the glass door of the food aperture, notes the level of alcohol in the 

 reservoir of the lamp, and at the exact moment when it reaches the 

 mark on the gage tube he closes the pinchcock at the bottom of the 

 burette. The alcohol supply bottle, with rubber stopper and glass 

 tubes, is then weighed, and the height of alcohol in the burette accu- 

 rately noted. At the end of the experiment the same operation is re- 

 peated, the lamp reservoir being again filled to the mark on the gage 

 and the level of the alcohol in the burette again recorded. The alco- 

 hol supply bottle is then weighed, and the difference in the weights 

 at the beginning and end, corrected for difference in the amounts of 

 alcohol in the burette, gives the quantity of alcohol burned. The 

 weighing of the bottle may be made with sufficient exactness on the 

 balance for weighing food, or that for weighing the water and carbon- 

 dioxide absorbers. 



During the course of the experiment, as the level of alcohol in the 

 alcohol lamp becomes low, sufficient alcohol may be admitted from 

 time to time to keep the level well above the lower end of the wick. 

 This successive addition of alcohol needs no special measurement, since 

 it is the total amount (loss in weight of the bottle) admitted during a 

 period that is actually required. Furthermore, there may be differ- 

 ences more or less great in the level of alcohol in the burette. It is 

 possible with care to adjust this amount to very nearly the same at the 

 end of each period by blowing over more or less alcohol from the bottle, 

 and this is regularly done, though differences in level of the burette are 

 invariably recorded and the residual amount of alcohol in the burette 

 allowed for in the calculations. 



On the particular burette used in connection with this lamp, the 

 graduations happened to correspond very closely indeed to the weight of 



