466 



SCIENCE 



[N. S. Vol. XXV. No. 638 



hybridization in poultry. DeVries's distinc- 

 tion between varietal and specific crosses, 

 based on their behavior in inheritance, does 

 not hold, as is shown by contrasting the be- 

 havior of those characters that are new to 

 the race with those that have been lost. 

 Furthermore a similar comparison shows that 

 deVries's view, that a patent character domi- 

 nates a latent one, as color does albinism, is 

 not always found to hold in poultry. Stand- 

 fuss's idea, that older characters or species 

 dominate, has been shown by Bateson and 

 Saunders not to hold good, and Davenport is 

 able to confirm their conclusion. Of nine- 

 teen characters examined nine old ones domi- 

 nated and ten new ones. The result depends 

 clearly on some physiological peculiarity of 

 the character that has no relation to its an- 

 tiquity. These and other conclusions of the 

 author are of importance for a fuller under- 

 standing of the laws of inheritance. 



This admirable study of Professor Daven- 

 port's gives promise of still more important 

 work to follow when the history of the ma- 

 terial is further worked out. The Carnegie 

 Institution is to be congratulated on the high 

 order of work accomplished under its auspices. 

 T. H. Morgan 



Columbia Univeksity 



SOCIETlEa AND ACADEMIES 



THE PHILOSOPHICAL SOCIETY OF WASHINGTON 



The 629th meeting was held on February 

 16, 1907, President Hayford in the chair. 

 The evening was devoted to a paper by Dr. 0. 

 G. Abbot, on ' Solar Radiation and Terrestrial 

 Temperature.' 



The speaker gave a general account of the 

 work of the Smithsonian Astrophysical Ob- 

 servatory in recent years, including its two 

 expeditions of 1905-1906 to Mt "Wilson. The 

 means of observing solar radiation both in 

 total, and for all parts of its spectrum, were 

 briefly described, and holographic spectrum 

 energy curves extending from wave-length 

 0.37 /A to 2.8 fi, were exhibited. Such curves 

 are obtained in twelve minutes each, and a 

 series of eight or ten, covering the hours 

 from low to high sun, is secured on each 

 favorable day. By means of the expression 



log (i=m log a -|- log d„ (in which d and d„ 

 are ordinates of such curves for a given wave- 

 length at the earth's surface and outside the 

 atmosphere, respectively, m, is secant of the 

 sun's zenith distance, and a the transmission 

 of the atmosphere for zenith sun) the form of 

 the spectrum energy curve outside the atmos- 

 phere is obtained. By means of diagrams it 

 was shown how closely this straight line equa- 

 tion fits the observations, and further evi- 

 dence of the trustworthiness of the formula 

 as a means of estimating the transmission of 

 solar radiation in the atmosphere is furnished 

 by the close agreement of simultaneous Wash- 

 ington and Mt. Wilson determinations of the 

 solar constant, though a mile difference of 

 altitude and 3,000 miles difference of longi- 

 tude is between the stations. 



About sixty values of the ' Solar Constant 

 of Eadiation,' as determined by Mt. Wilson 

 observations of 1905, were shown in a diagram, 

 and the evident change of nearly eight per 

 cent, in August, 1905, pointed out. It was 

 stated that single determinations were thought 

 to be accurate, relatively, to one per cent, for 

 usually good Mt. Wilson days. 



The average numerical value of the solar 

 constant is believed to be 2.12 calories per 

 square centimeter per minute; and high 

 values obtained by Langley, Angstrom, and 

 others, as quoted in text-books, can be shown 

 to be ill founded. Angstrom, indeed, has 

 himself withdrawn his value four calories, but 

 it still continues to be quoted. 



Observations of the reflecting power of 

 clouds were described and preliminary results 

 showing wide departure of clouds from ideal 

 matt surfaces were presented. Attention was 

 drawn to the very large amount of solar radia- 

 tion lost by reflection of clouds to space, and 

 the probability that variations of cloudiness 

 may in many instances mask the effect of 

 variations of the solar radiation. Inland sta- 

 tions, only, seem likely to yield trustworthy 

 evidence of direct connection between varia- 

 tions of the solar constant and the terrestrial 

 temperatures, but evidence was cited of a 

 direct connection of the kind as shown in Mr. 

 Langley's paper {Astrophysical Journal, June, 

 1904) where nearly one hundred stations over 



