^>20 Mr Stevenson Macadam on the 



would have risked the loss of iodine."* The other parts of the 

 arrangement need no comment. 



For six days the air unceasingly traversed the arrangement, and 

 at the conclusion not less than 100,000 cubic feet of elastic fluid had 

 passed through. I was unable to watch the experiment through- 

 out the entire period of its continuance, so that, after securely ar- 

 ranging the apparatus, and witnessing the commencement of its 

 action, I confided it to the charge of Mr John Begg, the intelligent 

 manager of the iron-works, who kindly took care of it till the close 

 of the period mentioned. I then dismantled the arrangement, and 

 transferred it to Edinburgh, where the results of the experiment 

 were ascertained. The gas bottle (1) was destitute of liquid. At 

 the lower part and around the sides saline matter in small quantity 

 was attached. On rinsing with distilled water, this was easily 

 washed out, and starch, hydrochloric acid, and nitrite of potassa, 

 were added to it. No iodine was present. The starched papers (2) 

 were not sensibly altered in tint. The contents of the bottles (3 and 

 4) were severally tested, as in the previous experiment, and no 

 iodine was present. 



From these results it was apparent, that in the large volume of 

 air subject to examination, there had not been an appreciable quan- 

 tity of iodine. Theoretically there is every probability of iodine 

 and bromine being present in the atmosphere ; the latter in much 

 greater quantity than the former; and it is only after such re- 

 peated failures that I have come to the conclusion that the quantity 

 of iodine in the atmosphere is frequently too minute for detection by 

 the ordinary methods of testing. 



The weather during each of the experiments was favourable to the 

 object I had in view. Several sunny days preceded each of the trials, 

 and in general the wind was north or north-east ; in other words, 

 blowing from the Frith of Forth landwards. The volume of air 

 experimented upon was in every case larger than that used by Chatin. 

 So far as I can gather from his papers, he employed 4000 litres of 

 air at Paris, which contained jj-^ui of a milligramme of iodine. f 

 This is equivalent to 880 gallons of air producing 30 o!ooo* u °^ a 

 grain of iodine, or aVoVV'i'*^ °f a grain of iodide of sodium. The 

 volume of air employed in the first unexceptionable experiment 

 which I made, viz., that where nitrate of silver was used, was 300 

 cubic feet, or about 1870 gallons, which, calculating from Chatin's 

 observation;;, and considering the delicacy of the test I used, ought 

 to have given satisfactory indications of iodine. In the first of 

 the Kinneil experiments, 4000 cubic feet, or 25,000 gallons of air 

 passed through the arrangement 1 , and this, according to the same 

 standard, ought to have given very distinct proof of the presence of 



' Gmelin'e Handbook of Chemistry, Wails Translation, vol. v., p. 248. 

 I Com p tee Rcmlus, tome xxxii., p. GG9. 



