April 25, 1913] 



SCIENCE 



645 



Belative Intensity of Prismatic and Grating Spec- 

 tra: J. S. Plaskett. 



The grating spectrograph used was briefly de- 

 scribed at the last meeting and is arranged to be 

 used in the Littrow form, giving linear dispersion 

 17.5 A per mm. and with incident and diffracted 

 pencils 30° apart giving 33.0 A per mm. A half 

 prism silvered on the back can be substituted for 

 the grating in the Littrow form, giving the same 

 dispersion at Sy. Comparisons of intensity were 

 made with the Ottawa three-prism and one-prism 

 spectrographs, giving practically the same disper- 

 sions at Hy. Spectra of the sun and of different 

 stars agree well in showing: (1) The grating does 

 not diffract more than 30 per cent, of the incident 

 light and the spectra are correspondingly weak. 



(2) The diffraction star spectra are practically 

 uniform in intensity between X 4800 and X 3850. 



(3) Prismatic spectra are relatively stronger in 

 tie blue and weaker in the violet than diffraction 

 spectra. (4) Diffraction spectra become equal in 

 intensity to the three-prism spectra at \ 4250, to 

 one and to half -prism spectra at \ 3970. Above 

 these regions prismatic are stronger, below weaker 

 than diffraction spectra. (5) The great loss of 

 light by absorption in prisms is shown by com- 

 parison of one- and three-prism spectra. The 

 former are more than twice as strong between Eg 

 and By, three times at X4250, seven times at 

 X4150, fifteen at Eg. 



A diffraction star spectrograph would be of 

 value in the ultra-violet, when spectra of uniform 

 intensity from E^ down were required, and in the 

 red end where prismatic spectra are unduly com- 

 pressed. 



A New Form of Cloclc Synchronization: E. Mel- 

 drum Stewart. ^ 

 The form of synchronization described is 

 adapted to the case where both the synchronized 

 and the synchronizing clocks control electric cir- 

 cuits. In the particular case where it is applied 

 at the Dominion Observatory the synchronizing 

 clock controls a circuit which is closed every 

 alternate second, while that controlled by the syn- 

 chronized clock is closed for one second every 

 minute, for the purpose of operating electric 

 " minute jumpers." Each of these circuits op- 

 erates a relay, and it is the coincidence of the 

 opening of the relays which forms the automatic 

 test of synchronism, which takes place every 

 minute. In addition there is used a neutrally 

 adjusted polar relay; the circuit from a local 

 battery is so arranged that, once a minute, whUe 



the relay operated by the controlled clock is closed, 

 current flows through the winding of the polar 

 relay, the direction of the current depending on 

 the position of the armature of the relay operated 

 by the synchronizing clock. Thus, at the instant 

 of the opening of the relay operated by the con- 

 trolled clock, the position of the armature of the 

 polar relay depends on whether the ' ' synchron- 

 izing ' ' relay is open or closed (that is, on whether 

 the controlled clock is slow or fast) ; and since 

 the polar relay is neutrally adjusted, it will re- 

 main in the same position until current next flows 

 through it, i. e., until the next even minute. As 

 soon as the relay operated by the controlled clock 

 has opened, its back contact is utilized, in series 

 with the points of the polar relay, to send a cur- 

 rent through one or other of two magnets in the 

 clock case, and so to either add to or remove from 

 the pendulum a small weight, so as to accelerate 

 or retard the clock for the following minute. At 

 the end of the minute the automatic comparison is 

 again made, and the clock again accelerated or 

 retarded as required. 



The controlled clocks are not particularly good 

 timekeepers, and are exposed to considerable vicis- 

 situdes of temperature; to ensure satisfaction the 

 correcting weights are made capable of taking 

 care of a variation in rate of 8 or 10 seconds per 

 day; the synchronization is in this case effective 

 to within about a hundredth of a second. In the 

 case of a high-grade clock very much smaller 

 correcting weights could be used, and the interval 

 between the automatic comparisons could be in- 

 creased to perhaps an hour. 



The principal advantage of this type of syn- 

 chronization is that there is no possibility of 

 stopping the controlled clock by interference with 

 the synchronizing current, an advantage which, 

 so far as I am aware, is not shared by any other 

 method. 



An Investigation of the 94-inch Photographic 

 Objective of the ShattucTc Observatory : H. T. 

 Stetson. 



The original 9.4-inch visual lens by Alvan Clark 

 has been made convertible into a photographic by 

 the substitution of a new flint, giving a focal 

 length of 10 feet 6 inches. Measurements of extra 

 focal plates taken after Hartmann's method for 

 determining aberration errors show extreme varia- 

 tion in the focus to be less than one part in 3,000 

 for the same wave-length. The greatest irregulari- 

 ties lie in zones of 55 cm. and 85 em. radii, where 



