18 DR. R. ANGUS SMITH ON THE 



show a steady diminution of oxygen, on one side, with occa- 

 sional risings as the wind may blow here or there, and a 

 steady rise of oxygen, on the other, with occasional fallings 

 also, as the wind may chance to carry smoke or other 

 gases. 



The laboratory stands in an open space, which contains 

 certainly a burial-ground in the centre, but is very much 

 freer from the smoke of the town than the streets are ; no 

 manufactories exist beyond or between it and the country. 

 The wind from west and south blows over many houses, 

 but over no large chimneys. 



The results clearly show a diflPerence between the air of 

 more and less pure places, and render the oxygen test more 

 valuable than it hitherto has been supposed to be. Un- 

 fortunately so many analyses are required, that the test 

 cannot be popular ; but as one to be resorted to when 

 the occasion warrants the labour it stands very clear. And 

 indeed how can it be otherwise ? We see putrid matter 

 laid on the ground, and find it disappearing rapidly, and 

 yet we are told that it is not accompanied by loss of 

 oxygen ; it is not credible, and the results given show it 

 to be incorrect. 



It may perhaps be said that, although some of the speci- 

 mens contain less than 20*9 oxygen, if they came from a 

 clear atmosphere, such air could not be considered very 

 bad. This reasoning cannot hold. Analyses are after all 

 subject to error, and the average is the only number on 

 which we can rely. It may even happen that the small 

 changes are caused by accidents which may give impurity. 

 For example, take gusts of impure air even in the air of a 

 street generally pure. 



It is abundantly clear that whenever we leave the region 

 of the uncontaminated or very little contaminated open air 

 we obtain a diminution of oxygen, although that diminu- 

 tion is very small ; this small loss is therefore a proof of 



