346 Prof. Gr. F. Armstrong. Diurnal Variation [Apr. 29, 



it was possible to ascertain whether any errors were being introduced 

 thereby. 



The jars used for collecting the air were the ordinary spice jars of 

 the confectioners, but of a somewhat unusually large size and with 

 mouths sufficiently wide to readily admit the hand for cleansing and 

 drying. The latter was effected by means of clean linen cloths specially 

 kept for the purpose. The mouths of the jars were easily and tightly 

 closed with Jennings' tin and india-rubber capsules. 



The jars employed were four in number, and were distinguished as 

 A, B, C and D. Their available capacities were as follow : — 



A= 14,260 cub. centims. B= 14,845 cub. centims. 

 C = 14,905 cub. centims. D=13,385 cub. centims. 



The quantities of air experimented upon are thus very considerable ; 

 indeed the jars are by far the largest hitherto employed in similar in- 

 vestigations since De Saussure used a globe of 34 litres ; while they 

 are nearly double the capacity of those used by Dr. Thorpe in his 

 Atlantic and Irish Sea experiments. 



The increased quantity of air thus taken for examination should, it 

 will be seen, tend to enhance the value and accuracy of the results, 

 since the possibility of detecting any differences that may exist in the 

 amounts of carbonic acid present in the air during the day and at 

 night is thereby greatly increased. 



The mode of filling the jars, the interiors of which were previously 

 carefully dried and polished, was partly by blowing the air into them 

 with a large pair of ordinary kitchen bellows and partly by violently 

 tossing them to and fro ; but a test experiment showed that the 

 latter means was alone sufficient for the purpose. 



Baryta water, the strength of which was frequently tested, was used 

 to absorb the carbonic acid, and 75 cub. centims. were introduced into 

 each jar. This quantity was chosen to allow of a duplicate titration 

 of 25 cub. centims. in each experiment. 



After being filled with air and the baryta water added, the jars 

 were well shaken and rolled on their sides so as to cover their internal 

 surfaces with the baryta solution, and they were then allowed to stand 

 not less than four hours. Several test experiments showed that this 

 time was sufficient to effect the maximum absorption of the carbonic 

 acid so attainable. 



Two jars at least were filled at each observation, and wherever the 

 results from one jar only are noted in the tables the omission is due to 

 some accident either in the filling, or at a subsequent stage of the 

 experiment. 



The exposure of the baryta water being completed, the contents of 

 each jar were quickly decanted into a stoppered bottle and allowed to 

 stand until the liquid was perfectly clear. And this was found to be a 



