108 EXPERIMENT STATION RECORD. 



Tlic author reviews and discusses the nii^thods of analysis in tlie 

 light of recent investigations, giving a hir^e number of references to 

 original papers. In connection with the protein determination ho rec- 

 ommends the separation of tiiis into albuminoids and non-albuminoid:: 

 by Stutzer's method, mentioning that in this division t!u' peptuncrsfall 

 in the group with the amids, etc., which is considered a sligh*: objec- 

 tion to the method. The fat determination should be made in dried 

 material with ether free from water and alcohol, and the residue dried 

 in a current of hydrogen or illuminating gas. The lecithin in the 

 extract should be dissolved out with hot absolute alcohol, tlie amount 

 calculated from the phosphorous content, and this deducted from the 

 ether extract. This gives an extract composed mostly of glycerids, 

 but containing also cholesterins and sometimes wax-like substances. 



The methods of F. Schulze, Hoftraeister, and Gr. Lange are mentioned 

 as available for thedeterminiitionof ''cellulose," but it is recommended 

 generally to determine " merely the crude fiber" by means of the "com- 

 mon well-known method." 



The starch is determined with the aid of malt extract. 



The constituents of the nitrogen-free extract soluble in malt extract 

 (starch) and in water (cane sugar, raf&nose, lupeose, etc.) are deter- 

 mined separately. The carbohydrates insoluble in water and malt 

 extract (gums yielding mannose, arabinose, xylose, etc.) may be aj^proxi- 

 mately determined by difference. 



The organic acids, when precipitable by lead acetate, can be approxi- 

 mately determined by adding lead acetate to the afiueous extract from 

 a known (piantity of material, filtering and washing, decomposing the 

 residue with hydrogen sulphide, and then determining the acidity ot 

 the solution by titrating with standard alkali. Deducting the sulphuric 

 and phosphoric acids gives the organic acids, which can not be calcu 

 lated to per cent unless the nature of the acids is known. 



In closing, the author calls attention to the fact that the methods at 

 hand for the analysis of seeds are far from perfect, and that the results 

 obtained by them must always be regarded as open to correction when 

 more Accurate methods are found. 



A simple method for the determination of fat in butter, W. S. 

 BwEETHF.n {PcnHsylvania h'ta. Rpt. If^QSyjjj). 143-140).— The method is 

 as follows : 



"The sample is melted aiirl then cooled as quickly as possible while being violently 

 shaken. Ten grams are melted and washed into a 200 cc. flask with 76 degree ben- 

 zin. After shaking well and cooling the flask is tilled to the mark with benzin. 

 The contents, after being well mixed, are allowed to settle, and 40 cc. is run from 

 a burette into a weighed platinum dish containing a few grams of asbestos. The 

 solvent is driven off by gentle heat, and the residue dried from 1 to 2 hours at 100- 

 C. and weighed. . . . 



"When the curd and ash are to be determined, the fat may be separated by wash- 

 ing it from the residue into the graduated flask by decantation. The residue is then 

 transferred to a Gooch crucible and manipulated as by the official method. A cheaper 

 grade of beuzin could probably be used as a solvent." 



