1042 EXPERIMENT STATION RECORD. 



This meter pump, whith w;is designed and built especially for this 

 apparatus by Mr. Frederick Hart, of Pou«hkeepsie, New York, acts 

 upon the same principle as the Blakeskn^ meter pump used in the 

 Atwater-Kosa apparatus, but differs materially from it in mechanical 

 details. A full description of it has l)eenpu))lislied." The pump con- 

 sists essentiall}" of two cylinders of drawn-steel tubing, 19 inches in 

 diameter, moving up and down in mercury. The pump is adjustable 

 to three different lejigths of stroke and four speeds. As thus far 

 used, it has been set to deliver approximately 50 liters per stroke 

 (exactly, 49.539) and has been run at the rate of al^out 14 strokes per 

 minute, the total ventilation, therefore, being about 700 liters per 

 minute. The number of strokes as recorded b}^ a revolution counter, 

 with the corrections for temperature and pressure, gives the total 

 volume of air passing through the apparatus, and the results of the 

 analysis of the ingoing air, calculated upon this volume, give the 

 weights of water and carbon dioxid cari-ied into the apparatus by 

 the current of air. 



By means of a shunt valve connected with an ingenious train of 

 gearing, one stroke is delivered at regular intervals alternately 

 through one or the other of two special outlets. The pump can be set 

 to deliver thus one stroke in 200, one in 400, or one in 800. The two 

 aliquot samples thus taken are conducted to two large pans having 

 counterpoised rubber covers substantially like those used in the 

 Atwater-Rosa apparatus. From these pans each sample separately is 

 aspirated by means of a subsidiary air pump and passes through a set 

 of six large U tubes (10^ inches), the first two containing pumice stone 

 saturated with sulphuric acid, the second two soda lime, and the last 

 two pumice stone and sulphuric acid. The increase in weight of these 

 tubes, of course, gives the amounts of water and carbon dioxid con- 

 tained in the samples, and this amount multiplied by the propei- factor 

 gives the total amount contained in the outcoming air. Subtracting 

 from this that contained in the ingoing air, determined as above 

 described, gives the amounts added b}" the animal. 



From the U tubes the air is conducted to the apparatus for deter- 

 mining the combustible gases excreted. This consists of a 1-inch 

 copper tube, having an effective length of about 60 inches, filled with 

 platinized kaolin and kept at a red heat by 64 gas-burners. In this 

 tu})e the combustible gases are oxidized to carbon dioxid and water, 

 which are absorbed and weighed in a second set of U tubes similar to 

 the first. Many difficulties were experienced in securing satisfactory 

 results with this part of the apparatus on account of the large volume 

 of air to be handled (3^ liters per minute as ordinarily used) and the 

 considerable force required to draw the air through the numerous 

 absorption tubes. The connections with the combustion tubes are 



«American Machinist, 25 (1902), p. 1297. 



