THE PURITY OF GLYCERIN. 29 



COMMENTS BY MANUFACTURERS. 



Replies were received from all of the manufacturers whose goods 

 were examined. The following excerpts from these letters are of 

 interest. This investigation shows that the glycerin available on the 

 American market is of excellent quality, the one sample which was 

 materially defective being of foreign origin. 



COLGATE & Co. 



We would call attention to the fact that the directions for Fehling's test as given 

 by the U. S. P. are not explicit enough, as it is not stated whether tie mixture is to 

 stand for six hours at ordinary temperature or on the water bath. The test is entirely 

 uncalled for, as any C. P. glycerin should leave but a very slight residue seldom 

 over 0.01 per cent on being allowed to burn off spontaneously, and if any sugar were 

 present it, as well as other nonvolatile impurities, would show in the amount of residue. 

 Silver tests, and ammoniacal silver tests, are unreliable, as it is easy to get results 

 ranging from no coloration to a heavy mirror of silver on the same glycerin by slight 

 variations in the method of testing. The test is no criterion for the presence of acro- 

 lein, as it is easy to work the test so as to give a heavy silver mirror with any glycerin 

 made, although acrolein may be entirely absent. I will add that the sample of 

 glycerin we sent you gave no appreciable reduction when we tested it with ammoniacal 

 silver nitrate U. S. P. 



The mixing of glycerin with an equal volume of concentrated sulphuric acid is 

 another test which is apt to give very erratic results. 



In reference to specific gravity it should be mentioned that the specific gravity given 

 in the U. S. P. as corresponding to 95 per cent glycerin is not correct within 1 per 

 cent of glycerin, but was calculated from old, inaccurate tables. Of the published 



20 f 1 

 tables, that of Nicol at ' is most accurate, being very nearly correct. 



ZO (j. 



Now that the U. S. P. has become the legal standard, it is desirable that the tests 

 should be rational in every particular, and that they should be so rigid as to exclude 

 inferior goods, and on the other hand should require no unnecessary tests and no tests 

 apt to give erratic results. 



Comments by the authors. The application of the designation 

 "C. P." to the available glycerin misrepresents the facts and should 

 therefore be discontinued. 



THE HARSHAW, FULLER & GOODWIN Co. 



Under date of May 18, 1907, we forwarded to you a sample of our glycerin, which 

 represented the average run of goods we were at that time turning out. This, how- 

 ever, was nearly two years ago, and since that time we have been constantly endeav- 

 oring to improve our product. We have succeeded so far as sulphate is concerned. 

 ^Your analysis of the sample referred to states that it showed a trace of sulphate. An 

 analysis of our present product would not show this. 



,- With reference to butyric acid, it seems impossible to entirely eliminate this. In 

 operating our glycerin refinery, we must purchase our crude material from a great 

 number of producers located in France, Italy, Spain, Germany, South America, and 

 elsewhere. No two of them make exactly the same product. There are certain pro- 

 ducers who make such an excellent quality of crude that it is possible for us to pro- 

 duce from it a refined glycerin without any objectionable features. The quantity of 

 this material, however, is comparatively small, and while we could send you a sample 



