Oct. II, 1877] 



NATURE 



505 



Weismann, the eminent zoologist, and Herr August 

 Gruber publish an interestinfc article on their joint 

 researches with regard to Daphniada, a family of Ento- 

 mostraca. These investigations were principally confined 

 to the species Aloiiia, of which M. rectirostris has been 

 well known for a long time. Weismann now describes 

 two new varieties and names them M. brachiata and M. 

 paradora. Other researches relate to the male forms of 

 the species Macrothrix and Pasithea. 



The Fishes of Lake Nicaragua.— Drs. Gill and 

 Bransford have recently investigated the fauna of this 

 lake and contributed a paper to the Philadelphia Academy 

 of Natural Sciences. The element of especial interest 

 is the association of characteristically marine forms with 

 fresh-water types. Thus, together with cichlids and 

 characinfe, none of which are marine, we have a species 

 of megalops, a shark, and a saw-fish. A similar combina- 

 tion occurs in the Philippines, where, in a fresh-water 

 lake, a saw-fish and a dog-fish are found. The megalops, 

 however, is not known elsewhere in fresh water so isolated 

 from the sea as Lake Nicaragua. These instances suggest 

 caution in generalising on physiographical conditions from 

 fossil remains. The most probable cause of such a com- 

 bination is the detention and survival of salt-water fishes 

 in inlets of the sea that have become isolated and 

 gradually transformed into fresh-water lakes. 



SOLAR RADIATION AND SUN-SPOTS 



In the year 1875 two articles by Mr. H. F. Blanford 

 on the connection between solar heat and sun-spots 

 appeared in the pages of Nature (vol. xii. pp. 147 and 

 18S), in which it was shown that Mr. Baxendell's con- 

 clusion that the sun's heat undergoes a periodical variation 

 coinciding directly with that of the spots, appeared to be 

 supported by the evidence of observations of the black- 

 bulb thermometer taken at certain stations in Bengal and 

 the neighbouring provinces. My attention has been 

 recalled to the subject by the almost complete failure of 

 the rainy monsoon this year in Upper India, and by the 

 excessively high temperature ever since the middle of 

 June, and I have been thereby led to attempt to discover 

 whether any evidence in favour of Mr. Baxendell's con- 

 clusion, or against it, is to be obtained from the registers 

 of meteorological stations in Upper India. I have there- 

 fore gone over the registers of certain stations in the 

 North- West Provinces and Oudh, where solar radiation 

 temperatures have been recorded since 1869, and at which 

 not more than one change of instrument occurred in the 

 interval 1869-1876. The inference I draw from these 

 records is exactly the opposite of Mr. Blanford's. They 

 do not afford any support to Mr. Baxendell's theory, but 

 the energy of solar radiation appears from them to be 

 most intense when the spots are fewest. 



The reason of this discrepancy in the two sets of results 

 will probably be found in the diffeient modes of treat- 

 ment we have adopted in extracting from the registers 

 their evidence regarding this question. Mr. Blanford's 

 treatment of the Silchar register, the results of which are 

 given in his first paper, consisted in picking out certain 

 " clear days " on which the mean serenity at 10 A.M. and 

 4 P.M. was 6-ioths of the expanse, or more, tabulating the 

 maximum temperatures of solar radiation for all these 

 days in each month, and taking the average, neglecting 

 the months of the south-west monsoon which are almost 

 entirely wanting in clear days, as above defined. The 

 results of the examination of the Darjiling register, 

 given in the second paper, were obtained by deduct- 

 ing from the three highest recorded temperatures of 

 solar radiation in each half-month the corresponding 

 maximum temperatures in shade, tabulating these differ- 

 ences for each month, and takmg the average. Both 

 these devices, I think, introduce new elements of error, 



probably as great as those they were intended to obviate ; 

 for, as Mr. Blanford himself points out, it constantly 

 happens that the solar radiation thermometer record's 

 much higher temperatures when the sky is partly covered 

 with broken cloud than when it is perfectly clear, the 

 reason being, doubtless, that the instrument is then 

 screened to a great extent from radiation into space, 

 while the sun comes out from time to time, and exerts his 

 full heating power upon it. I am therefore incUned to 

 think that the somev/hat higher radiation temperatures 

 recorded at the Bengal stations in 1870, 1S71, and 1872 

 were probably due to the larger number of partially 

 cloudy days in those years as compared with the years 

 immediately preceding and succeeding them. 



In the clear atmosphere of Upper India the months 

 of March, April, May, October, and November are gene- 

 rally almost without a cloud, the mean serenity at stations 

 on the plains during those months being over 7-ioths. 

 There can, therefore be little error in taking the mean 

 excess of the maximum temperature of solar radiation 

 above the maximum in shade during those months in 

 each year, as the measure of the intensity of solar radia- 

 tion during the year ; for the two irregularities introduced 

 by occasional cloudy days, prevention of direct radiation 

 from the sun to the thermometer and prevention of radia- 

 tion from the latter into space, will to some extent 

 counterbalance each other. 



The following table gives these yearly means for three 

 stations : — Chakrdta, lat. 30° 40' N., long 77° 55' E., 

 elevation above sea-level, 7050 feet ; Roorkee, lat. 29° 52' 

 N., long. 77'' 56' E., elevation, 890 feet ; and Lucknow, 

 lat. 26° 50' N., long. 81° o' E., elevation 370 feet. 



1869 1870 1S71 1872 1873 1874 1875 1876 



Chakrata ... 607 572 627 612 630 dyi^ tyd 6^-6 

 Roorkee ... 51-9 390 41-5 474 540 522 51-6 55-9 

 Lucknow ... 44-5 43-5 472 47-6 470 47-6 49-1 54-2 



Mean ... 52-4 466 50-5 52-1 547 55-1 56-1 582 



It will be seen that the lowest of these numbers is that 

 corresponding to the year of sun-spot maximum, 1870, 

 and the highest is that for 1S76, a year of very few spots. 



Unfortunately none of the thermometers in use at these 

 stations between 1869 and 1876 had been compared either 

 with a standard or with the others, and as the instruments 

 at some of the stations were in the meantime replaced, 

 these results are doubtful within the limits of the error of 

 such thermometers. This error probably never exceeds 

 5°, but the difference between the numbers for 1870 and 

 1876, given in the table, amounts, on the average, to 1 1-6°. 

 It is believed, too, that during the whole period, 1869-76, 

 the thermometer at Chakrata was never changed, and one 

 thermometer was in continuous use at Roorkee from 1872 

 to 1876. The differences in the table must therefore be 

 the effect either of a real variation in the sun's heat, or of 

 a greater degree of absorption than usual during the wet 

 and cloudy years, about the sun-spot minimum. Lest 

 they should be attributed to this latter cause, I have 

 examined the registers of the same three stations in much 

 the same way as Mr. Blanford did that of Silchar, and 

 find that owing to the proportion of cloud being so very 

 small, especially in October and November, the results 

 are very little changed. The monhs of the south-west 

 monsoon, June, July, August, and September, and those 

 of the winter rains, December, January, and February, 

 have been left out, and the only days counted during the 

 remaining months are those on which the cloud proportion 

 at four P.M., when the maximum thermometers were read, 

 did not exceed 2-ioths. The only exception to this rule 

 is the hill station Chakrata, where the cloud proportion 

 for the spring months had to be fixed at one-half. The 

 exact temperature of solar radiation thus determined 

 varies also in the way shown above, the only apparent 

 effect of the treatment being the introduction of slight 



