ORGANIC MATTER 0¥ HUMAN BREATH. 235 



large glass flask^ surrounded by ice_, or snow and salt, by 

 which a temperature of several degrees below zero was 

 obtained. 



In the first essays, the number of breaths was counted, 

 and the flask washed out with distilled water; but this 

 was soon found to be unsatisfactory, as the extent of the 

 expirations varied so greatly. 



The aqueous vapour was then collected, and measured 

 and tested as follows : — 



If enough fluid had been obtained, a certain quantity 

 (say Kixxx) was mixed with 50 cub. centims. of distilled 

 water and tested for free ammonia by means of the Nessler 

 test. 



An equal portion of the fluid was then mixed with 30 IT]^ 

 of a saturated solution of carbonate of soda, and about 

 10 oz. of pure distilled water, ascertained by distillation to 

 be free from ammonia. 



The mixture was then distilled, and the distillate was 

 tested for ammonia until it ceased to give any indication 

 of its presence. 



This testing would give all the free ammonia, together 

 with any of this gas arising from the action of the car- 

 bonate of soda — for instance, from the decomposition of 

 urea (see Water Analysis, p. 55). 



50 cub. centims. of a strong solution of permanganate of 

 potash and caustic potash was then added to the retort, 

 and distillation again continued ; the quarntity of ammonia 

 now given ofi* would arise from the destruction of organic 

 matter. 



The results of these examinations are given in the 

 Tables at the end (pp. 245-247), Table I. giving the 

 records of healthy breath. Table II. of breath from persons 

 affected by various disorders. 



In both Tables are given, in successive columns, (i) the 

 number of the observation, (2) the nature of the case. 



