J34 ^- Lindhard. 



tuations of the barometer in my experimental series are considerably 

 greater than are ordinarily met with under our conditions. 



In the experiments of February 1909 the carbonic acid tension 

 in the first half of the series follows very exactly the rise and fall 

 of the barometer; this part of the series also shows greater regularity 

 in other regards than the remaining part; but the same feature is 

 also recognizable in the latter in spite of a couple of inversions, and 

 if we split up the series by grouping together the determinations at 

 a barometric height of 762 ^ or less into one group, the remainder 

 into another, we obtain an average for the first group of 34- 1 and 

 for the second 343, a difference which is certainly not considerable 

 but is nevertheless when taken in conjunction with the remaining 

 series worthy of note. The temperature in the room was compara- 

 tively uniform during the whole period of experimentation. 



The series of experiments for April 07 is different. In these 

 experiments we see an interplay of opposing forces, which make 

 the results difficult of interpretation; but as shown by the distribu- 

 tion of the deviations indicated on p. 105, the л^аг1а11оп8 here also 

 move on the whole parallel with the fluctuations of the barometer;' 

 and if we divide up the series, so that the experiments where the 

 barometer is > 760 mm. are taken in one group, we obtain a very 

 sharp separation, which even considering the temperature conditions 

 can scarcely be explained without assuming some influence of the 

 air-pressure. For the barjameter < 760 the average is 300, when the 

 barometer is > 760 the average is 31-8. 



In the series for June, on the other hand, there is no parallelism 

 in the variations from day to day; dividing up the series, for a 

 barometric pressure > 764 the average is 2965, when the barometer 

 is < 764 the average is 2865. It is to be remarked here, however, 

 that as mentioned earlier some large deviations occur, which on 

 dividing up the series tend to place a disproportionate weight on 

 the side on which they fall. 



The August experiments, excluding the last experiment in which 

 there is a decided effect of the cold, show no tendency to group 

 themselves according to the air-pressure. And this holds good like- 

 wise for the November experiments, where the daily variations 

 sometimes go in the one direction, sometimes in the other. A 

 grouping of the experiments here, however, gives a slightly lower 

 value for the carbonic acid tension in the portion where the air- 

 pressure is highest. The difference is however quite inconsiderable. 



In the winter experiments of January 1908 the carbonic acid 



^ The value chosen so as to have the 2 groups somewhat of the same size. 



