The observations of MM. Boyé and Cailleux are of considerable interest. In conjunction with those of Ahimann and Seligman they give good evidence of the petrographic similarity between the ice at the ends of alpine glaciers and that of the marginal zones of ice caps which have been subjected to summer melting.

Nevertheless, it seems to me, as it does to the authors, that further observations are very desirable before one can definitely fix the characteristic sizes of the glacier grains in these regions.

Further, the slight heterogeneity of these sizes poses a delicate problem, for the interpretation of the problem, not of the sizes, for which two alternate hypotheses can be stated:

1. a. Homogeneity is the last phase of the growth of the crystals (by coalescence or recrystallization or both these factors).

2. b. Homogeneity results by change following melting, the small crystals disappearing before the larger ones.

The total mass of the small grains is relatively small, and when these grains melt they cannot have great influence upon the “index of smoothness“of the larger grains.

For temperate glaciers, as for the marginal ice studied by MM. Boyé and Cailleux, the exclusive interpretation of the homogeneity by the first alternative cannot be accepted without reservation. The final phase of the growth of the crystals must be studied well inside the body of the glacier to the exclusion of any process of change. Nevertheless, the interesting work of these authors, and the reserve with which they temper the discussion, shows once again the necessity to direct the study of the physics of glaciers towards the conditions in the interior of the ice mass which so far have been studied all too little.