36 



SCIENCE 



[N. S. Vol. XXXVII. No. 940 



for the arithmetical mean of all, 20".526 ± 0".0057. 

 There are, however, very good reasons for assign- 

 ing higher weights to some of the 22 values than 

 to others. The results have been derived from 

 observations at two different places with prac- 

 tically four different instruments. We find, how- 

 ever, for the weighted means practically the same 

 value as that given above, 20".525 ± 0".0045. The 

 corresponding value of the solar parallax is 

 8".7S0.- 



A Test of the ISi-inch Objective of the Dearborn 

 Observatory Telescope: Philip Fox. 

 It has been interesting to apply Hartmann 's test 

 to this objective, the high quality of which is 

 established by the long series of difficult double- 

 stars which have been discovered with its aid by 

 both Bumham and Hough. The perforated screen 

 contained 96 holes, 12 mm. in diameter, arranged 

 on 24 zones. The resulting curve of zonal foci is 

 very smooth. The extreme range in focal length 

 for the different zones is about 2 mm. The mean 

 of the two sets gives for the Hartmann criterion 

 T, the result 2"= 0.30. 



Some Besults from the Personal Equation Appa- 

 ratus of the 9-inch Transit Circle of the V. S. 

 Naval Observatory: W. S. Eichelbebger and 



F. B. LiTTELL. 



With a personal equation apparatus the follow- 

 ing results were obtained for the three observers, 

 L, M, P, positive corrections indicating that the 

 observer anticipates. 



CHEONOGRAPHIC OBSERVATIONS 



EYE AND EAR OBSERVATIONS 



Assuming that the personal equation can be 

 represented by p -\- m sec S ± n see S, the coeffi- 

 cients were computed and corrections applied, thus 

 reducing the errors to a very small quantity, gen- 

 erally less than 0^.01 for ehronographic, and less 

 than 0^.05 for eye and ear observations. 



* See A. J., 639, and Pub. Flower Observatory, 

 Vol. III., 2. 



Comparisons were made to test the reliability 

 of these results, with satisfactory consequence. 



A Sigh-level Measurement of Solar Badiation: 



Feank W. Very. 



Applying the methods elaborated in a previous 

 paper, ' ' The Violle Actinometer as an Instrument 

 of Precision, " " to the reduction of an observa- 

 tion by M. Violle, who obtained from a sounding- 

 balloon record at an altitude of 13,700 m. an 

 excess of 53° in sunshine above an air temperature 

 of 65° below zero Centigrade, the author obtained 

 for the observed solar radiation-equivalent at 

 13,700 m., 2.86 gram cal./sq. cm. min. Making 

 allowance for the absorbent atmosphere, deriving 

 the value from the consideration of observations 

 at sea-level, Keeler's and Nanry's on Mt. Whit- 

 ney, Violle 's on Mt. Blanc, the value of the solar 

 constant obtained is 3.5 gram cal./sq. cm. min. 



A Criterion of Accuracy in Measurements of 

 Atmospheric Transmission of Solar Padiation: 

 F. W. Very. 



The coefficient of atmospheric transmission of 

 solar radiation on a given day is variable because 

 of the variation of atmospheric extinction with 

 zenith distance, and because this absorbent quality 

 changes in the middle of the day, especially if 

 much moisture is present. Thus in Bouguer's 

 formula, 



there are two variables p and e, and the equation 

 is soluble only by approximation. 



To accomplish this approximation, M. Crova 

 has proposed that the curve of radiation shall be 

 considered as the envelope of a series of loga- 

 rithmic arcs. By constructing successive sub- 

 tangents to this envelope at points having values 

 of e for abscissae, coefficients of transmission of 

 the sifted rays, at stages corresponding to deter- 

 mined values of e, can be found by the equation 



T=i 



«a+e). 



where 



-(m:)--(t)' 



This procedure usually gives a series of varying 

 values of m, so that a solution of the problem is 

 no nearer, unless some occasion can be found 

 when m does not vary appreciably. Such an 

 extraordinary occasion is on record in the observa- 

 tion by SavSlief at Kief, Eussia, December 28, 



' Pub. A. A. S. A., 2, 90. 



