Analysis of the Human Breath. 1 09 



chamber ") t and was arranged in the same manner as that figured 

 in the Philosophical Magazine for September 1861, and described 

 by Professor Tyndall in his memoirs " On the Absorption and 

 Radiation of Heat by Gaseous Matter/' The source of heat was 

 a small flame of carbonic oxide. It is well known in what a 

 capricious and unstable manner carbonic oxide burns in the open 

 air, as it is extremely difficult to keep a small flame of this gas 

 from being extinguished when unguarded from the currents of 

 air present in a room. The first consideration was therefore to 

 provide a means of protecting the source, so as to bring it under 

 control and render it at once steady and uniform. The follow- 

 ing method was finally adopted and adhered to throughout the 

 investigation. 



A glass globe 4 inches in diameter had its upper part drawn 

 into a short chimney, and its lower into a neck which fitted 

 tightly into a brass gallery, this being held at right angles by a 

 brass stem. At a short distance from the gallery the stem was 

 bent perpendicularly downwards, and then passed into a suitable 

 stand in which it could be raised or lowered at pleasure, thus 

 enabling the lamp to be fixed at any desired elevation. The 

 globe was pierced by a third opening midway between the chim- 

 ney and the neck. Through this aperture, which formed a tubu- 

 lure 1*1 inch in diameter, the radiation from the source passed 

 unchanged in quality into the experimental tube. Part of the 

 interior surface of the globe opposite the central opening was 

 thickly coated with silver-leaf, which by reflexion increased the 

 total radiation without sensibly altering its character. Through 

 the neck of the lamp passed a brass tube terminating in a small 

 jet, by means of which any gas could be conveyed and burnt in 

 the centre of the globe. 



Carbonic oxide gas, contained in a large air-holder, was then 

 forced by gentle pressure through a regulator, and finally caused 

 to issue from the burner fixed in the lamp. The gas burnt with 

 a small blue flame about half an inch in length, which could be 

 adjusted by means of stopcocks or by slightly altering the regu- 

 lator. And thus a source of heat nearly constant from day to day, 

 and extremely steady, was obtained. In front of the lamp was 

 mounted the brass experimental tube, 49*4 inches long and 24 

 inches in diameter, its ends being stopped by polished plates of 

 rock-salt. As the diameter of the tube was greater than that of 

 the opening in the lamp, a diaphragm was made of double 

 polished brass having a central aperture of the same size as the 

 opening. When fitted to the end of the tube next the source, 

 the stop completely suppressed all radiation from the heated 

 sides of the glass globe. 



The radiation from the flame, after passing through the expe- 

 rimental tube, fell on the anterior face of a thermo-electric pile, 



