52 



Mr. J. B. N. Hennessey. 



[May 25,. 



then the question arises, for a given radiation effect how far hack 

 must we seek for the corresponding spot impulse (if any) ? But my 

 present view being necessarily limited to what appears on C 2, I 

 advance (in imagination) the area curve by some three days, bringing,, 

 for instance, its features of 12th under that of radiation for 15th ; 

 when this is done, there is more general resemblance between the two 

 curves, indicating increased radiation for increased spot-area. The 

 middle or highest rising portions (16th to 29th) of spot-area will now 

 correspond fairly with 18th (or 19th) to 1st of radiation ; an interval 

 about the same as a semi-solar rotation. 



18. Looking at the radiation curve (autumn 1881) alone, it may be 

 described as consisting of three hills (of which the last, alas, is cut 

 short with the series and leads to nothing) and intervening valleys ; 

 between the apices of the first and second hills we have eleven days,, 

 between the valleys six and fourteen days. 



19. This brings me to T. 9, where are tabulated the means of such 

 of the autumnal results as happen to be forthcoming, lying between 

 15th October and 16th November for each year, and also the summer 

 mean for 1881. Now, looking at the yearly autumn means reduced to 

 45° Z.D., it will be seen that they exhibit a continuous decrement of 

 about 2 per cent, annually from 1879 to 1881. I think that they also 

 have the appearance of truth about them, so far as they go, and, in fact 

 in obtaining them considerable care and exactness was exercised, so that 

 they may claim reliance. But when we look at the summer mean of 

 1881, which in my judgment is equally reliable, we find it only 565, 

 at a solar zenith distance of but 15° '5, or nearly 30° less than the 

 autumn Z.D. Unfortunately, I repeat, no long series could be taken 

 in the summer, but the autumn long series in the same year (1881) 

 o-ives a mean decrease of 47 radiation for an increase in Z.D. of 17° 

 reckoned from Z.D. 46° ; and accepting this as some approximate 

 basis for reduction of the summer result from 15°'5 to 45°, I can only 

 conjecture that this summer result, if thus reduced, would be less than 

 the autumn result of the same by, say, at least, 40 to 50 units, or 

 about half a degree F.* This represents so great and unexpected a 

 change that one naturally looks for flaws in the summer observations, 

 but I can find none. The mean is obtained from five days' observa- 

 tions taken carefully in (visibly speaking) unimpeachable weather, 

 and the five results are accordant (see T. 7). Beyond this I can only 

 say that I will certainly try to obtain other daily series, and a long 

 series, in the approaching summer. 



* The solar negatives for the summer of 1881 are not with me now, but judging 

 from silver prints in my possession, it is certain that sun-spots and faculae were 

 very sensibly absent in summer 1881 compared with sun-spot exhibits in the suc- 

 ceeding autumn ; if this be accepted as cause, variation of radiation with sun-spot 

 area wouhd receive very decided confirmation in the present instance. 



