254 



NA TURE 



[JULY 12, 1906 



STAR TRANSITS BY PHOTOGRAPHY.'' 

 T^HE annoyance that arises from the effects of a " magni- 

 -'■ tude equation " in transit observations has led to 

 various suggestions for its detection or removal. Screens 

 in front of the object-glass so as to reduce the light of 

 bright stars have been employed with advantage, and 

 various photographic devices arranged with the view of 

 eliminating personal peculiarities have been adopted. But 

 while ingenuity has been active in proposing practical 

 applications and methods, the numerical results have been 

 few. Recently, Prof. S. Hirayama, of the astronomical 

 observatory at Tokyo, has put in practice a contrivance 

 similar to that employed by the Rev. Father Hagen in 

 photographing a star in the focus of the transit telescope. 

 In this method the exposure and occultation of a star i^ 

 alternately effected by means of a bar, moved in obedience 

 to a clock, so as to give rise to a series of dots along the 

 star trail. 



The Tokyo transit was for this purpose provided with a 

 triple object-glass, reducing the secondary spectrum, and 

 specially corrected for photographic rays. The aperture 

 was 13-5 cm., and the focal length 211 cm. The range 

 of magnitude to which the telescope was applicable de- 

 pended, of course, upon the time of exposure permitted by 

 the occulting bar. As a matter of fact, with a full second's 

 exposure, equatorial stars of the fifth magnitude gave a 

 measurable image. For stars of greater declination than 

 73° the exposure of one second was too short to divide 

 distinctly the successive impressions from each other. The 

 limitations of the method are thus clearly indicated. F^or 

 fainter stars it seems necessary to consider the possi- 

 bility of moving the photographic plate at the same rale 

 as the star, and imprinting on the plate the image of a 

 fixed reticule at known times. The simpler method adopted 

 by Prof. Hirayama recommended itself to him, since the 

 apparatus could be constructed in the workshops belonging 

 to the observatory. 



This apparatus consisted of a camera containing the 

 reticule, occulting bar, and the dark slide, which could be 

 inserted in the place ordinarily occupied by the wires and 

 eye-piece. The reticule consists of seven fine lines ruled 

 upon a microscope cover-glass, firmly cemented to a rect- 

 angular frame which carries the dark slide. These lines 

 are interrupted for a short distance in the middle of the 

 field so that they shall not interfere with the star images. 

 The centre of the field is marked by two horizontal wires 

 in the ordinary manner. The occulting bar (Fig. i) is a 



-Showing the ba 



, in the 



; of the field. 



thin metal slip about 8 cm. long with a square opening 

 at one end, so as to allow the observer to see the stai 

 enter the field, and to permit him to adjust the instrument 

 so that the transit shall take place behind the bar when in 

 its stationary position. The end of this bar is soldered to the 

 armature of an electromagnetic coil. Whenever the electric 

 circuit is established the bar is lifted up and the star 

 exposed. This circuit is made and broken automatically 

 by contact springs in the standard sidereal clock. The bar 



] " Preliminary Experiments on the Photographic Transit." By S. 

 Hirayama. Annales de I'Observatoire astronomique de Tokyo. Tome iii., 

 4^ fascicule. (Tokyo, 1905.) 



consequently operates as an exposing shutter, permitting 

 the cone of light from the star to fall for a longer or 

 shorter period upon the sensitised plate, the period • being 

 decided by the contact springs. 



The sensitive plate when inserted in the dark slide 

 comes within 0-2 min. of the lines of the reticule, so that 

 these lines and the image of the star are practically in the 

 same focus. Evidently this distance must be made as 

 small as possible to reduce any error arising from photo- 

 graphic parallax, but the plate can be shifted in its own 

 plane, so that five separate exposures can be made upon 

 ilie same plate. The advantage thus secured cf taking 



ed, sho 



^lilly enl.-irged. 



NO. IQI5, VOL. 74] 



five stars on the same plate is somewhat discounted by 

 the fact that no proper adjustment can be made for de- 

 veloping the plates according to the different actinic 

 intensity of the stars. 



The method of observation will be easily apprehended 

 from the description of the apparatus and the character 

 of the results obtained (Fig. 2). The measurement of the 

 plates is not so simple. It is distinctly admitted that to 

 measure a negative is more laborious than to read the 

 fractions of a second from a chronographic sheet. 

 Theoretically, the beginning of each " break " made by 

 the clock is the e.xact point to which the reading should 

 refer ; but owing to the difficulty of measuring the edges 

 of the dots, due to the want of sharp definition, this plan 

 could not be adopted. The middle of the " break " to 

 the middle of the " make " has been taken as the full 

 second. This arrangeinent, or conventional rule, has prob- 

 ably got over the difficulty arising from the photographic 

 spread, for it seems not impossible but that the want of 

 definition at the edges of the dot, or the distance between 

 two dots, is dependent upon the brightness of the star. 

 But if this source of error is eliminated the author has tO' 

 regret that the length of the dot depends upon the battery, 

 the spring, the friction, and the moving parts of the 

 apparatus as affected by the variable component of the force 

 of gravity. " The weakening of the battery has been con- 

 stantly provided for, but at present I see no way of escape 

 froiTi all the other disadvantages." 



This admission seems to deprive this peculiar method 

 of observation of much practical benefit. The question that 

 has to be solved is not so much one of relative accuracy 

 as it is of the possibility of eliminating systematic errors, 

 inherent in older and more familiar methods. Looked at 

 as a simple matter of determining the position of a star 

 on a plate at any required moment, the results leave nothing 

 to be desired. In an example worked out in detail it is 

 shown that the error in a single pair of measures is 

 + 0017s., and the mean error of thirty-two pairs, or what 

 may be regarded as equivalent to a complete transit, 

 + 0-002S. The results of the measures of 140 stars, made 

 when the plate was moved with, and against, the direc- 

 tion of diurnal motion, gave for the average value of 

 personal equation -I-0027S., the positive sign implying that 

 the time of transit was longer when the plates were 

 measured along the diurnal motion than when measured 

 against it. 



But such measures leave the question of a possible error 

 dependent upon magnitude untouched. Unfortunately, the 

 limited range of magnitude and the small nuinber of 

 observations do not permit any very definite conclusion to 

 be drawn. The author presents a table of forty-six stars 

 in which the photographic magnitude varies from 1-2 mag. 

 to 5.5 mag., and gives residuals for each night and the 

 mean residual. The latter is less than 005s. in all but two 

 cases out of the forty-six. Further, when these mean 

 residuals are arranged for each star in the order of photo- 

 graphic magnitude, no relation between the two is notice- 



