256 Scientific Proceedings, Royal Dublin Society. 



The following arrangement has been lately adopted by myself, 

 and, both as regards simplicity and efficacy, leaves nothing to be 

 desired. It is, in fact, so simple that, to a superficial observer, 

 my communication might at first sight appear superfluous, but 

 the importance of the question involved renders any apology 

 unnecessary.* 



At ordinary temperatures very weak solutions of ammoniacal 

 gas (such as we have to deal with in the analysis of potable 

 waters) are fairly permanent; but, in the gaseous conditions, 

 owing to the different tensions of the vapour of water and 

 ammonia, this stability no longer exists. Again, small traces of 

 ammonia in the gaseous condition are instantly absorbed by 

 water in the non-gaseous condition. These are the two equal 

 sources of error to be guarded against in the distillation of 

 ammonia from potable waters, and, bearing these two points in 

 view, I adopted the arrangement figured below. 



It consists of a retort fitting into a fairly long-necked receiver, 

 with an india-rubber stopper. The arrangement must be air- 

 tight. The receiver is connected with a bent tube proceeding 

 from its stopper, and connecting it with two bulb tubes of a 

 special form, and marked respectively " A " and " Z." This 

 marking is to prevent any confusion in the hurry of manipulation, 

 as they are exactly the same shape. The bulb tubes are some- 

 what similar to a flat Liebig's potash bulb, but with two pear- 

 shaped bulbs on each side, to prevent regurgitation of the fluid, 

 and three absorbtion bulbs at the bottom. The centre bulb of 



* One of the recent important works upon chemistry, is " Roscoe and Schor- 

 lenmaers' Treatise." In their description of the determination of ammonia in potable 

 waters, there is no provision against this fruitful source of error. If the determination 

 were carried on as it is figured at page 252, Vol. I., of that work, the results would be 

 utterly unreliable in any laboratory where other analyses were being performed. 



