620 ANNUAL REPORT SMITHSONIAN INSTITUTION, 1911. 



Table I. — Lethal dose. 

 [January to March. 1910.] 



On breathing two to three parts per million, we ourselves find it 

 irritating to the respiratory tract, with a tendency to produce, in this 

 concentration, headache and oppression. The irritation set up by 

 ozone, together with its strong characteristic smell, affords ample 

 warning, and would prevent anyone exposing himself unintentionally 

 to a dangerous concentration. The irritation set up would naturally 

 make anyone remove himself from the influence of the ozone before 

 any serious damage to the lungs had been set up. As far as we can 

 see, then, no serious risk can arise from the use of ozone generators 

 so long as the generators are not placed in a confined space from which 

 escape is impossible. 



It is only j)ossible to estimate concentrations of much less than one 

 part of ozone })er million parts of air by passing very large quantities 

 of the ozonized air through the acidified potassium iodide solution. 

 We find concentrations of far less than one in a million parts can be 

 both smelled and tasted; the physiological test for ozone therefore is 

 extraordinarih r delicate. If ozone is used in a ventilating system, 

 we think it should be in such concentration as is scarcely perceptible 

 to a keen sense of smell. 



Ozone has most potent action as a deodorizer. We tested this by 

 filling our experimental chamber with the smoke of shag tobacco, 

 ammonium sulphide, or carbon bisulphide vapor. At other times 

 we placed in the chamber stinking meat, or human feces. After 

 putting in action the two small ozonizers, placed in the roof of the 

 chamber, for two minutes, we were not able to detect the odors of 



