October 20, 1911] 



SCIENCE 



529 



rated as 10 on an absolute scale of intensity, and 

 the strength of any line in absolute units is the 

 product of its width by the intensity of its absorp- 

 tion, multiplied by a constant. The constant may 

 be chosen so that unity coincides with Rowland's 

 unity, and it is found that in this case the above 

 definition is very closely fulfilled, although Row- 

 land's numbers diverge considerably in other parts 

 of the scale. 



From a smooth curve which represents the true 

 photometric intensities of ultra-violet lines on 

 negatives by Higgs, the significance of Rowland's 

 symbols has been found to be as follows: 



These values have been confirmed by measures 

 in the green. 



It is shown that, in general, the relation between 

 the width and the intensity of the photographic 

 images of spectral lines is not as simple as has 

 been assumed. It is also shown that differences 

 of width of spectral lines have been exaggerated 

 in Thollon's Atlas, doubtless in order to convey a 

 more realistic impression of the intensities. 

 Periods of Variable Stars in the Small Magellanic 

 Cloud: Henrietta S. Leavitt. 

 (Read by Professor Russell.) 

 The periods of 25 of the variable stars in the 

 Small Magellanic Cloud have been determined, 

 and have been found to vary in length from 

 1.25336 days to 127.0 days, while the light curves 

 are of the chister, sometimes called the ' ' ant- 

 algol" type. The brightness and the length of 

 the period are so closely related that if one is 

 known, the approximate value of the other may be 

 inferred. For an increase of one magnitude in 

 brightness at maximum or minimum, the logarithm 

 of the period increases by about 0.48. It seems 

 possible that all these variables are of similar 

 mass, those whose periods are long having slight 

 densities, and vice versa. On account of their 

 faintness, it is impracticable to obtain their spec- 

 tra, with our present facilities. A number of 

 brighter variables having similar light curves, as 

 UY Gygni, are in isolated positions, and should 

 repay careful study not only of their spectra, but, 

 if possible, of their parallaxes. Among other 

 questions suggested are the following: 



1. Are there definite limits to the mass of vari- 

 able stars of the cluster type? 



2. Does the spectra of such variables having 

 long periods differ from those of variables whose 

 periods are short, and how? 

 The Sadial Velocities of 96 EereuUs: S. A. 



Mitchell. 



Preliminary results from the measures of 

 twenty-five spectrograms of 96 Hereulis taken 

 with the one-prism Bruce spectrograph of the 

 Yerkes Observatory show the presence of four 

 components and a period of 50.2 days. On ac- 

 count of the large velocities, which range on the 

 plates measured from — 98 kilometers per second 

 to + 74 kilometers per second, it was possible to 

 measure three separate components on half the 

 plates, and four components on a few of them. 

 The components all belong to the B-type. 



This interesting system will be followed closely, 

 and more definite results will be given when more 

 plates are obtained. 



The San Luis Observatory of the Carnegie Insti- 

 tution: R. H. Tucker. 



The first part of the expedition sailed from New 

 York in August, 1908. 



Ground was broken for the observatory, on land 

 set aside by the national government, a month 

 later. 



The piers were built of reinforced concrete, and 

 the observatory and dwelling of brick, with wooden 

 roofs covered with rubberoid. The shutters for 

 the meridian circle building were constructed at 

 Albany, before starting, and were taken out by 

 freight along with the rails for the floor, instru- 

 ments for preliminary adjustment, chronometer 

 and one of the astronomical clocks. 



Five months later the dwelling was ready for 

 the staff of ten men, and the observatory was 

 completed for the installation of the Pistor and 

 Martins instrument of the Dudley Observatory. 



Observing on the regular list was begim in 

 April, fundamental and miscellaneous observing 

 being taken up. The full list contained 15,000 

 stars. About 1,600 of these were fundamentals, 

 of varying degrees of precision; from 4 to 32 

 observations were taken of each; usually the ob- 

 servations have been in excess of the requirements. 

 The first year 61,000 observations were made, 

 though, from the time observations were in full 

 swing, 62,000 were made in one year. About 300 

 nights were available for observations, of which 

 200 were clear throughout, in the first year. At 

 the close of the work, in January, 1911, there had 

 been made 87,000. 



