Death in Closed Vessels 511 



and replace it by mercury, which then rises into the bowl C. If 

 then the cock R is closed (position 1) and the reservoir B is set 

 at the bottom of the groove, the mercury will fall in the tube and 

 remain at 76 cm. above the level of reservoir B; in other words, 

 there will be a barometric vacuum in chamber A. If then cock R 

 is turned (position 3) to put this chamber in communication with 

 the lateral tube which, in our figure, communicates with a system 

 of sleeve and balloon serving only for the extraction of blood gases, 

 a certain quantity of outside air is introduced, and the mercury 

 descends in the barometric tube. When the cock is closed (position 

 1), a quantity of this air is imprisoned; and if any is needed for an 

 analysis, one needs only to raise reservoir B and shift cock R to 

 position 2, and the air driven out by the rising mercury passes 

 through the little bowl C and enters the inverted tube which is 

 ready to receive it. 



The invention of the mercury pump is usually attributed to 

 German technicians, and with the love of foreign advertising cus- 

 tomary to us we often decorate this instrument with the name of 

 "Geissler pump." The truth is that the invention belongs in prin- 

 ciple to M. Regnault. Long ago this famous professor of the Col- 

 lege de France invented a similar pump, equipped with a three-way 

 cock, which is the most important part of the instrument. But 

 instead of using a movable reservoir, since at that time rubber 

 was seldom used in the construction of- apparatuses, he put his 

 barometric tube in communication with two reservoirs, one above 

 and one below; this required, of course, a rather complicated sys- 

 tem of tubes and cocks. But the principle was the same, and the 

 addition of a rubber tube is certainly not of sufficient importance 

 to make us forget the real inventor. 



And now it is very simple to understand how this instrument 

 can be applied to extracting the air contained in the bell-jars of 

 the apparatus represented in Figure 15. The only thing necessary 

 is to add to the lateral tube, which in Figure 16 communicates with 

 the balloon D, a rubber tube which can sustain a vacuum, and 

 which covers with its other end the cock M, which is at the top 

 of the bell- jar in which the experimental animal is contained. 

 When this cock has been closed and a vacuum has been made in 

 chamber A, I put cock R in position 3 so as to exhaust the air from 

 the rubber tube S; then close cock R (position 1), raise reservoir 

 B as high as possible, place the cock in position 2, and the air 

 passes out over the mercury in the bowl, over which the tube has 

 not yet been inverted. When this procedure has been repeated 



