624 



SCIENCE 



[Vol. LV, No. 1432 



ence for any month does not exceed 1° F. 

 Comparing corresponding slopes of the two 

 watersheds, the mean temperature is substan- 

 tially the same. (2) Precipitation occurring 

 as rain is practically equal on both watersheds. 

 If the soil is saturated, as small a rain as 0.01 

 inch may cause the streamflow to respond; but 

 ordinarily rains of 0.10 inch or less in summer 

 merely replenish losses due to evaporation or 

 transpiration, and do not affect streamflow ap- 

 preciably. Most of the summer rains are not 

 in excess of 0.25 inch, hence it is seen that 

 summer rains are not, in general, of great 

 importance. (3) A little less than 50 per 

 cent, of the precipitation is snow, but it yields 

 more than half the runoff. The average depth 

 of snow per season is 113.3 inches. The maxi- 

 mum observed was 149.7 inches and the mini- 

 mum 80.7 inches. 



Interesting features of the streamflow records 

 are: (1) Stream ^-1 rises more rapidly than B 

 and reaches a maximum sooner than B, but 

 before the flood has subsided a secondary max- 

 imum with a steadier flow may occur at A. 

 This feature, as mentioned above, is easily ex- 

 plained by topography. (2) Winter and 

 autimm show very little diurnal variation of 

 streamflow; summer is more marked, with a 

 maximum in the early morning hours and a 

 minimum between 1 and 2 o'eloek in the after- 

 noon; spring, however, with the great amount 

 of melting snow, has a pronounced diurnal 

 period owing to alternate freezing and thawing. 

 The amplitude of variation is greater at A than 

 at B; and the A maximum and minimum are 

 more pronounced. (3) An estimated disposi- 

 tion of 21.00 inches of precipitation, the aver- 

 age annual amount for eight years observations, 

 is shown for A as follows : 



Evaporatiou 7.39 inches 



Transpiration 3.91 inches 



Interception 3.62 inches 



Runoff 6.08 inches 



Total 21.00 inches 



It is clear that the objective of all these 

 studies is an accurate estimate of the relations 

 between the various factors on A and B in 

 order that, in the years following denudation, 

 the conditions on A can be used as an index 



to what would have occurred on B had denuda- 

 tion not been effected. It is only in this way 

 that the effect of the presence or absence of 

 trees can be ascertained. Much of the paper, 

 therefore, is devoted to these relations in too 

 great detail for abstracting. Thirteen "rules" 

 are developed as statements of these relations 

 to be used in the later discussions. These con- 

 cern ratios of discharges in different periods 

 and at different times, time intervals between' 

 crests, probable height of crests, and the depo- 

 sition of silt. 



This experiment is of great practical im- 

 portance with respect to hydrologieal prob- 

 lems — floods, irrigation, etc., and its outcome 

 will doubtless be watched with the greatest 

 interest by those who are concerned with these 

 problems. 



C. LeRoy Meisinger 



Washington, D. C. 



SPECIAL ARTICLES 



AN EARLY STAGE OF THE FREE-MARTIN 



AND THE PARALLEL HISTORY OF 



THE INTERSTITIAL CELLS 



The theorj' that the intersexual condition 

 of the free-martin depends upon hormones 

 secreted by interstitial cells of the testis of the 

 male twin and distributed by its blood to the 

 female 'depends primarily upon the demon- 

 strated connection between foetal vascular 

 anastomosis and the intersexual condition of 

 the female twinned with a male calf, and sec- 

 ondarily on comparative data. The time of 

 effective action of the male hormone has been 

 presumed to be very shortly after the beginning 

 of sex-differentiation in the embryo (Lillie, 

 '17) owing to the known normal condition of 

 the embryonic membranes in such stages, which 

 I'enders vascular connection possible, and the 

 very profouiid nature of the effect. The earli- 

 est stage of the free-martin hitherto described 

 is 7.5 em greatest length (Lillie, '17; Chapin, 

 '17). Sex-dift'erentiaition begins at approxi- 

 mately 2.5 cm. The gap thus indicated in our 

 knowledge of this phenomenon is now largely 

 filled up by study of a free-martin of 3.75 cm 

 greatest length, and of the complete history of 

 the interstitial cells of the testis and ovary 

 from 2.5 cm throughout life. 



