187^.] Various Altitudes on the Island and Peak of Jeneriffe. 507 



mixed with 100 cub. centims. of distilled water gave, in order to 

 neutralize 5 cub. centims. of the oxalic acid solution, 9'00 cub. 

 centims. as the mean of six determinations. My normal solution of 

 barium, analysed in London before leaving for Teneriffe, had yielded 

 8*92 cub. centims. ; the difference was only by 0*08 cub. centim. This 

 result would give a very slight deficiency of carbonic acid, but the 

 error might be expected to correct itself in a number of experiments. 



Finally, it might be objected that, in my Alpine experiment, a loss 

 of carbonic acid had been experienced from the india-rubber bag into 

 which the expired air was collected. In these experiments a certain 

 time elapsed after filling the bag previous to the air it contained 

 being introduced into the tube ; this lapse of time ranged between a 

 few minutes and thirty-five or forty minutes, and was required to 

 allow the air in the bag to cool down to the temperature of the water 

 in the tube. No doubt, after a certain time, an escape of carbonic 

 acid might be expected to take place through the substance of the 

 india-rubber bag, but no such escape, in any appreciable degree, could 

 have occurred during the above-mentioned period. This I determined 

 experimentally by subjecting a sample of expired air to analysis 

 immediately after filling the bag, and another sample of the air from 

 the same bag some time later. The results from four analyses made at 

 Cannes in February (1879) were as follows : — 



Experiment. 



CO2 expired 

 per minute. 



Time bag was 

 exposed to the 

 air. 



C0 2 found 

 after waiting. 



Difference per 

 cent. 



1 



0-451 



25 minutes 



0-454 



"66 more. 



2 



0-411 



27 „ 



0-407 



-97 less. 



3 



0-382 



50 „ 



0-380 



52 „ 



4 



-462 



30 „ 



0-469 



1'5 more. 



It is, therefore, obvious that, in my experiments on the Alps, no 

 appreciable loss of carbonic acid through the substance of the bag 

 took place previous to the air being subjected to analysis.* 



* In the whole of these experiments the air had been aspired into the tube for 

 analysis by means either of a nearly saturated solution of common salt or of water. 

 It had not occurred to me, at first, that the fluid adhering to the inside of the tube- 

 would have a material influence on the volumetric analysis which was to follow. 

 I determined the mean volume of fluid thus left in the tube, from 14 experiments, 

 to amount to 3 - 3 cub. centims. ; an error thus crept into the analysis, which, though 

 not interfering with the results as to the carbonic acid expired in the Alps, relatively 

 to each other, had, however, to be corrected when these results were compared with 

 those obtained at Teneriffe. It was calculated for every experiment separately 

 both in the Alps and at Teneriffr, and the correction was made accordingly. This 

 work proved very laborious, and delayed considerably the completion of this paper. 

 There is another probable slight source of error to be noticed in the analysis con- 



