U. S. DEPAETMEISTT OF AGRICULTURE. 



lation of the original composition of the materials from the analytical 

 figures would be practically out of the question in all but a relatively 

 few and simple cases. 



Even where chemical analysis, either qualitative or quantitative, 

 gives the desired information as to the existence of a certain com- 

 pound, the time consumed in making the analysis is comparatively 

 long. A method which would give reliable and rapid identifications 

 is obviously desirable and for certain work imperative. For instance, 

 a chemical analysis would readily give the percentage of the various 

 elements in a mixed fertilizer, but further data as to the actual com- 

 pounds of potash, say, which are present m the sample might be 

 necessary for certain plot or pot experiments. 



A number of methods designed to furnish evidence as to the con- 

 stituents of mixtures have been proposed and used in certain limited 

 fields of study. The separation of minerals in rock powders according 

 to their specific gravities by means of solutions of appropriate densities 

 was long in vogue, and even now is used occasionally. Ordinary 

 microscopicgil observation of crystal forms has been used extensively, 

 and, though sometimes abused, in the hands of an expert it can be 

 made of considerable value. Its application, however, is limited to 

 well developed crystals, and even then it furnishes only meager data 

 of a definitive kind. The danger attending its use becomes apparent 

 when one tries to identify one of the numerous ''needle-shaped" 

 crystals with no more data than can be obtained with the cross hairs 

 and possibly the rotating stage of the microscope. Microchemistry 

 at its best gives scarcely better results than a qualitative chemical 

 analysis. The observation of absorption spectra with a micro- 

 spectroscope is of extremely Rmited application. Up to the present 

 time, petrographic methods are the best and apparently the only 

 general solution of the problem. 



PETROGRAPHIC METHODS. 



The possible range of application of petrographic methods to chem- 

 ical problems is almost unlimited. They have been successfully 

 applied to the identification of compounds formed in concrete and 

 cement mixtures and to general ceramic problems. It has been 

 found possible to differentiate between phrenosin and kerosin derived 

 from both the brain and the soil. The alkaloids and sugars afford 

 notable instances of the rapid identification of substances without 

 tedious chemical procedures.^ Most phase rule work presents 

 problems easily solved by the petrographic microscope. This appli- 

 cation has already found commercial use in the study of brines. In 

 the laboratory of the Bureau of Soils it has been shown to be feasible 



» Wherry, E. T., U. S. Dept. Agr., Bui. 679 (1918). 



