586 JOURNAL OF FORESTRY 



nected with a true determination of snow-fall, or of the amount of 

 snow available for streamflow and irrigation at the beginning of the 

 melting season. Very wisely, the writer refrains from any broad con- 

 clusions. No one is more competent than Mr. Church to realize the 

 difficulties in the way of accurate snow measurements ; he has both the 

 practical knowledge to surmount these difficulties and the ability to 

 express himself clearly in writing. 



The present paper by Church is not of as great interest to foresters 

 as the earlier paper which I have already cited. In fact, it is almost 

 wholly given up to the technique of snow measurements and cannot be 

 said to present any results beyond those of special interest to engineers, 

 and perhaps foresters, who may attempt snow surveys. Briefly, the 

 present situation is : 



1. That the measurement of snow-catch, in gages of any kind so far 

 devised, is very inaccurate and gives no approximate idea, for high and 

 wind-exposed mountain regions, of the snow available in the spring, 

 which may contribute to irrigation. Every factor tends to make the 

 snow-catch less than the actual fall. (At Wagon Wheel Gap there have 

 usually been measured on the ground, in spite of evaporation through- 

 out the winter, water equivalents over considerable areas in excess of 

 the total recorded precipitation, and this where the gages are only ex- 

 posed to mild winds.) 



2. Snow surveys, to be accurate enough for streamflow prognostica- 

 tions, must include an enormous number of measurements, so arranged 

 in fixed places or in survey lines as to obtain a true representation of 

 the dift'erent topographical features of the basin, the open and forested 

 ground, etc. 



3. The Mount Rose snow sampler, a tube of 1% inches outside diam- 

 eter and of whatever length may be necessary to reach the maximum 

 depths of snow, has been mechanically perfected, so that under all con- 

 ditions of snow and ice samples are obtained satisfactorily and with a 

 small probable error from loss (crowding out to the sides) or accretion. 

 Such apparatus is always most successful in the hands of its designer, 

 but only because the latter understands its eccentricities better and uses 

 the apparatus more carefully than others. We believe, from the great 

 variety of conditions under which this sampler has been used satisfac- 

 torily, it can fairly be said that the mechanical problem has been solved. 

 Universal use of such a piece of apparatus is desirable, but, on the 

 other hand, the designing and use of new apparatus often brings to 

 light new phases of the problem under study. 



4. Although mere measurements of snow depth, if sufficiently nu- 



