15 



The general opinion of the collaborators is that the constant stirring 

 method of obtaining the titer is the most satisfactory and that the 

 method of saponitication is immaterial. The chief point of nonagree- 

 ment is on the diying of the fatty acids. 



CONCLUSIONS. 



In conclusion, it seems to the referee that the work of this year has 

 brought out the following points: 



(1) The method of preparing the fatty acids for the test is immate- 

 rial, and the analyst should be allowed to use the method which seems 

 most convenient to him. There is some danger from fire in alcoholic 

 saponification, but this method is quicker and there is less likelihood 

 of incomplete saponification, though, as Mory has shown, this question 

 is not a serious one. A number of chemists have used methyl alcohol 

 and find it equally satisfactory and much cheaper. The alcohol should 

 be completely removed where the alcoholic method of saponification 

 is used. The scheme given in the Dalican method effects this removal 

 completely. 



(2) The fatty acids should be dry. It seems to the referee that a 

 careful study of the results of this year will prove that the method 

 outlined in the Dalican scheme, i. e. , filtering and diying twenty minutes 

 at 100 C. , is sufficient and convenient, but does not change the titer. 



(3) The method of stirring is the chief factor in the difference in 

 titer, and extraordinary care should be taken by the operator to follow 

 exactly the method prescribed in order that comparable results ma} 7 be 

 obtained. 



(4) The Wolf bauer method of stirring gives higher results, especially 

 on fats of a low titer; it gives more uniform results where there is 

 a variation in temperature; it is the more scientific, and gives more 

 nearly the true titer. 



(5) There should be a standard thermometer for titer tests, so that 

 all thermometers used for this purpose would have the same size bulb, 

 the same length of stem, and be made of the same kind of glass, and 

 thus a stem correction could be accurately applied, or else the ther- 

 mometers should be standardized under the conditions under which 

 they are used. The latter would seem to be the more satisfactory 

 plan. 



PROPOSED METHOD. 



The -associate referee offers the following tentative method for criti- 

 cism and discussion, hoping that there will be a full expression of 

 opinion : 



Weigh 75 grams of fat into a metal dish and saponify by using 60 cc of 30 per cent 

 sodium hydrate (36 Baume" caustic soda) and 75 cc of 95 per cent by volume alco- 

 hol, or 120 cc of water. Boil down to dryness, with constant stirring, to prevent 

 scorching. This should be done over a very low flame or over an iron or asbestos 



