726 G. SHERBURNE ROGERS 
washing or floatation may be used, thus removing the finely 
divided and thoroughly hydrated—perhaps colloidal—material. If 
more elaborate methods are desired, recourse may be had to the 
elutriator, the electro-magnet, heavy liquids, etc.1 These methods 
of course destroy the proportions of the constituents, but facilitate 
the study of the features—assortment, angularity, and polish,— 
which are really important to the consideration of the rock from a 
genetic standpoint, and they will in many cases be the most expedi- 
ent. On the other hand it seems probable that certain very fine- 
grained shales and clays are largely made up of colloidal material, 
which may prove to be homogeneous in composition and to have 
more or less distinct optical properties, and in such cases it may be 
desirable to study it as a whole without attempting at first to 
ascertain the character of the mineral particles from which it was 
derived. 
Secondly, the question of field criteria for recognizing the 
particular rocks likely to prove critical presents itself, since it is 
obviously impossible to make and examine a complete collection of 
all the variations in a series. It is patent, of course, that very 
fine-grained rocks, since they do not respond satisfactorily to our 
present elementary methods of study, are less likely to prove 
helpful than the coarser-grained varieties; and on the other hand 
that the presence of uncommon constituents in very coarse sand- 
stones and conglomerates can usually be recognized without the 
microscope. The writer must confess his inability to suggest any 
definite rule, however, and believes that from the nature of the case 
none can be framed. Such a question must be left largely to the 
field geologist. 
Finally, it seems desirable that the geologist who studies the 
rocks in the field should be the one to examine them microscopically. 
The petrology of a sedimentary rock is in most cases of little or no 
significance except when considered in relation to the structural and 
stratigraphic features. Furthermore, its exact horizon is generally 
important and the relation of the stratum to those adjoining; and 
such data will, of course, be noted by the collector in any case. 
1See F. H. Hatch and R. H. Rastall, Petrology of the Sedimentary Rocks, 1913, 
Appendix on the Systematic Examination of Loose Detrital Sediments, by T. Crook. 
