730 



SCIENCE 



[N. S. Vol. XXX. No. 777 



ther improvement of making one of these supports 

 adjustable and determining its best position by 

 actual experiment. 

 The Focal Curves of the Single-material Camera 



Doublet of the Single-prism Spectrograph of the 



Detroit Observatory : R. H. Cuetiss. 



This paper dealt with a recent investigation of 

 the focal curves from X 3900 to X 6000 of the nevr 

 camera lens mentioned in the title. The collim- 

 ator is an Isokumat of 27.5 inches focus and 

 1.4 inches aperture. The camera lens has a focus 

 of 16 inches. Nine different combinations of 

 collimator and prism settings were tried covering 

 all cases that might be advantageous. It was 

 found that the deviation from straightness of the 

 focal curves was practically the same in all cases 

 over a distance of 34 mm. But for the portion of 

 the curve corresponding to the interval X 4000 to 

 X 5900 the deviations were least for minimum 

 deviation settings in the neighborhood of X 4400. 

 It was found that the entire region from X 3900 

 to X 6000 could be photographed in sharp focus 

 upon one negative. 

 The 'New Spectrograph Measuring Engine of the 



Detroit Observatory : R. H. Cuetiss. 



This engine was constructed from designs by 

 Dr. Curtiss based upon his experience with instru- 

 ments of the Zeiss, Toepfer and Gaertner types 

 as well as with one made by the John A. Brashear 

 Company for the Allegheny Observatory after 

 designs by Professor Schlesinger. The principal 

 features are: a sector for inclining the engine at 

 any desired angle; a long clock spring for taking 

 up the back-lash of the screw; a reversible sec- 

 ondary plate carriage; a motion of rotation of the 

 microscope around an axis parallel to the screw; 

 the use of an interrupted reticle and a removable 

 reticle holder to carry glass reticles of any type. 

 The screw of the engine was made at the observa- 

 tory shop and seems to possess a remarkable 

 accuracy. 

 Solar Spectroscopic Observations: Phiup Fox. 



Results in three lines of work were presented: 

 ( 1 ) Spectrograms of a dark calcium flocculus that 

 had a high velocity. (2) A brief report on the 

 work undertaken by Dr. Abetti and the author on 

 the sun-spot spectrum. The photographs were 

 obtained with an 18-foot Littrow spectrograph 

 used with an horizontal telescope of 60 feet focal 

 length. The investigation covers the region from 

 X 3900 to X 6800 and includes about 8,000 altered 

 lines. (3) A preliminary report concerning a 

 comparative study of the spectra of the details 

 of the photospheric granulation. 



The Use of Quartz Fibers for Micrometer Wires: 



Philip Fox. 



Spider-threads, while excellent in many ways, 

 have two disadvantages: they are affected by 

 humidity and they are too coarse. It is a difficult 

 matter to find spider threads whose diameters do 

 not exceed the resolving power of the telescope. 

 The author has made some experiments with 

 quartz fibers furnished by Professor Nichols. 

 There was no difficulty in finding and mounting 

 fibers that were one third of the usual diameter 

 of spider threads. They are very smooth and do 

 not readily retain dust particles, they are easily 

 illuminated and are not afi'ected by changes in 

 humidity. One fiber has been in use for more 

 than a year. 



In the discussion that followed this paper at- 

 tention was called to the fact that owing to the 

 non-elastic qualities of the fiber there was a tend- 

 ency for it to work loose from the mounting, 

 under the temperature conditions that usually 

 prevail at a telescope. 



Some Dynamical Considerations on Globular Star 



Clusters: P. R. Moulton. 



The assumption on which this paper was based 

 is that the dimensions, masses and relative veloci- 

 ties in the globular star clusters are such that 

 they maintain essentially constant dimensions. 

 Consequently, if this assumption is sound, when 

 two of the three classes of data are furnished by 

 observations, the third, e. g., the masses, is given 

 by the equational relation which is developed. 



If TT is the parallax of the cluster, D the diam- 

 eter of the cluster in the same units, W the num- 

 ber of stars in it, then p, the average distance 

 between adjacent stars, is 

 D 



P = 



fNn 



If y = 5,000, D = 30', 7r = 0".01, which is the 

 mean parallax of fifth magnitude stars according 

 to Kapteyn's formula, we have p = 10,500 astro- 

 nomical units. 



An important question in the discussion is 

 whether a star passing through the cluster has 

 many close approaches to other stars, and whether 

 it may be prevented from leaving the cluster by 

 the general gravitative control of the whole group 

 of stars. If R is the radius of the cluster in 

 astronomical units, the probability, P, that a star 

 passing through the cluster will pass within r 

 astronomical units of at least one other star is 



-K0- 



