April 19, 1907] 



SCIENCE 



611 



from the D lines to \ 3,300 in the violet, 

 a distance of 9.5 inches. 



Measurements are now being made on 

 the three or four thousand lines of the 

 flash spectra. These photographs show a 

 dispersion about the same as the Bruce 

 three-prism spectrograph of the Yerkea 

 Observatory and the Mills spectrograph of 

 the Lick Observatory, or a dispersion one 

 fifth that of a twenty-one-foot grating of 

 ordinary Rowland mounting in the first 

 order spectrum, or one tenth of the second 

 order. The eclipse spectra were practically 

 normal. Wave-lengths thus far deduced 

 show a probable error less than five-hun- 

 dredths of an Angstrom-unit. 



Comparison of Results of Oiservations 

 with the Reflex Zenith Tube and Zenith 

 Telescope at the Flower Observatory 

 during 1905 and 1906: C. L. Doolittle. 



The Temperature of 

 Lowell. 



Mars: Percival 



On the Absence of Long Heat Waves in 

 the Sun's Spectrum: E. F. Nichols. 



Formulas for the Comparison of Astronom- 

 ical Photographs: Haeold Jacobt. 

 This paper contains formulas suitable 

 for the direct comparison of rectangular 

 coordinates measured on different astro- 

 nomical negatives. The problem here in- 

 volved supplements what may be called the 

 fundamental transformations in the reduc- 

 tion of celestial photographs; viz., the cal- 

 culation of right-ascensions and declina- 

 tions from rectangular coordinates, and 

 rectangular coordinates from right-ascen- 

 sions and declinations. 



"Light Curves of New Variable Stars of the 

 Algol Type, and of Short Period: Hbnei- 



ETTA S. LeAVITT. 



In a recent circular of the Harvard Ob- 

 servatory thirty-six new variable stars were 

 announced, mostly brighter than 9.5 at 



maximum. All are within fifteen degrees 

 of B. A. 12'', Dec. — 60°, the region in- 

 cluding both the Southern Cross and the 

 Nebula in Carina. About sixty new vari- 

 ables in this area, mostly fainter than the 

 tenth magnitude at maximum, had been 

 announced in previous circulars. A large 

 proportion of the new objects appear to 

 have short periods, and several are of the 

 Algol type. These are now being studied 

 with a view to determining their periods 

 and light curves. The periods of four 

 have already been announced, those of 

 three others are now made public for the 

 first time. All but one of the seven vari- 

 ables are of the Algol type. The periods 

 vary from nine tenths of a day, to five and 

 one third days, while the smallest and the 

 largest ranges observed are four tenths of 

 a magnitude, and two and one half magni- 

 tudes, respectively. 



The variable C. P. D.— 50° 3,809 is of 

 special interest, as the period is equally 

 divided by a secondary minimum. The 

 principal minimum is nine tenths of a 

 magnitude fainter than the normal bright- 

 ness, 9.3, while the secondary minimum is 

 four tenths of a magnitude fainter than 

 the normal. The average deviation from 

 the light curve of a single observation is 

 very large, being no less than ± 0.14 

 magn., while the average deviation for all 

 the other Algol variables here described, is 

 only ± 0.07 magn. This large deviation 

 does not appear to be due to accidental 

 errors of observation, though the variable 

 was difficult to measure on many of the 

 plates. It has not been found possible to 

 improve the period on the supposition that 

 it is constant, and it is probably to be cor- 

 rected by a third term, not yet determined. 



C. P. D.— 49° 6,972 has a large range. 

 The faintest magnitude observed is 11.5, 

 which is two and a half magnitudes fainter 

 than the normal brightness; but no obser- 

 vation at minimum has as yet been seciired, 



