154 BACON. 



accuracy in this determinatioiij yet I consider that this disadvantage is 

 more than overbalanced by its simplicity. 



The present method of determining the ammonia coefficient in the 

 urine is quite beyond the ordinary practitioner. A high ammonia index 

 of the urine has been i\sed in cases of pernicious vomiting of preg- 

 nancy as a safe diagnostic indication that the pregnancy should be 

 terminated. It has been shown very clearly that in many cases a delay 

 of a comparatively few hours in operating involved fatal results. Wlien 

 a patient's life is at stake, it is highly important to have a quick method 

 for making this determination which can be cari-ied out by the phy- 

 sician in his own laboratory. It is also probable that there may be 

 wide variations in the ammonia index of the urine in many pathologic 

 conditions. At the present time this determination is but rarely made 

 on urines, consequently but little is known concerning the number of 

 disturbances in the human organism which may cause a large part of the 

 nitrogen to be eliminated as ammonium salts instead of as urea. 



I have sometimes obtained ammonia indices varying as much as 10 per cent 

 from the calculated, in several hundred determinations which I have made to test 

 the method outlined in this paper, using known quantities of urea and ammonium 

 chloride: thus with a true ammonia index of 4 there may be obtained 3.6 to 4.4 

 per cent. This is a very large error in an analytical determination, but neverthe- 

 less at the present time, for diagnostic purposes, an ammonia index of 22 would 

 probably have the same significance as one of 24. When more accurate deter- 

 minations become necessary this method must be improved. 



The proposed method is based upon two well-known reactions. When 

 urea is treated -^vitli Millon's reagent (mercuric nitrate in nitric acid 

 solution), it is decomposed to give nitrogen and carbon dioxide and this 

 same reagent has no effect on ammonia or ammonium salts. 



As is well Imovm, hypobromites in alkaline solution decompose urea 

 into carbon dioxide and nitrogen, the carbon dioxide being absorded ; they 

 also decompose ammonia, liberating nitrogen. The reactions may be 

 represented by the following equations : 



In acid solution ivith mercuric nitrate. — 



CO(]SrH2)„-f30 — > C02-fN„+2H20 



In alkaline solution with hypolromites. 



C0(NH„)„+30 — > CO„+N.-f 2H2O. 



2NH3+36 — > N2+3H.O. 



CO(]SrH2)2.+3NaOBr+2NaOH — > 3NaBr+Na„C03+]sr2+3H20 



2NH3+31SraOBr — > N„+3NaBr-f SH^O. 



The gas evolved by reason of the action of urea on an alkaline hypo- 

 bromite solution is nitrogen. Both nitrogen and carbon dioxide are 

 given off by the action of urea on the acid mercuric nitrate solution 

 and as they are evolved in equal volumes, half of the gas obtained is 



