162 Mr. J. B. N. Hennessey. [Dec. 16, 



Table VIII, there are obvious inconsistencies in these hourly results, 

 such, however, as have been experienced on some other occasions 

 with similar actinometers. No doubt, in point of eligibility as a 

 station, Mussooree is very far superior to Dehra, as already shown, 

 so that small fluctuations may be expected to disappear at the latter, 

 through the dissipation of energy exacted by its denser envelope ; 

 but the sums of the defects present large measures of comparative 

 radiation before and after noon, and while Mussooree gives a decided 

 preponderance of radiation to the former period, this does not hold at 

 Dehra, as may be thus shown : — ■ 



Nov. 12. Nov. 14. 

 Sums for the four hours before noon ......... 346 .... 432 



afternoon ... .' 346 .... 420 



After noon — Before noon ......... .... 12 



Mean (indicating excess of radiation before noon) .... 6 



This discrepancy in evidence may, however, be fully accounted for 

 by local haze, which is certainly absent at Mussooree, and as un- 

 deniably is present at Dehra ; accepting this conclusion, the observa- 

 tions show a fact not without interest, viz., that at the last-named 

 station the local haze absorbs more heat relatively in the forenoon 

 than in the afternoon, a conclusion which can be supported by other 

 exhibits from the same table. And, as regards the inconsistencies in 

 question, whereby the radiation in some instances decreases with 

 decrease of zenith distance, these, also, may to some extent arise from 

 passing local causes, such as are inevitable in the neighbourhood of a 

 large town. But after all these points have been considered, there 

 still remains the fact that separate "casts-on" of the instrument are 

 likely to be burdened each with a constant error, by which the whole 

 series may be affected. 



16. But though ineligible in the form of group means (as given in 

 Table VI) to test the conclusion (1) of Article 14, the evidence of the 

 Dehra results may be legitimately employed when presented in the 

 form given in Table VII, where the mean results are for + half- 

 hourly time-ranges. Thus we find 



Before After Eefore — 



noon. noon. After. 



Mussooree mean of first eight days .... 954 .... 947 .... 7 

 ,, „ remaining seven days 966 956 .... 10 



Dehra mean of complete seven days .... 882 .... 880 .... 2 



which are in accord and in keeping with the hypothesis advanced as 

 to dissipation of energy. 



