884 



SCIENCE 



[N. S. Vol. XXXII. No. 833 



however, it differs radically as regards the criteria 

 for distinguishing rising and falling temperatures. 



Results from, Photographic Photometry : Edwabd 

 S. King. 



Early in the work the photometric laws were 

 tested photographically; for example, the law of 

 the square of the distance from the source of 

 light was confirmed photographically by using 

 various apertures. The law of the cosine for 

 oblique rays of light was shown to be photo- 

 graphically valid up to inclinations of 60°. Be- 

 yond that point the intensity became less, possibly 

 because of roughness of the film. The measures 

 have all been made with a photographic wedge. 

 Usually, three settings are made on each image; 

 these are made at intervals some time apart. The 

 average deviation of the wedge measures is about 

 ± 0.05 magn. The wedge is capable of measuring 

 small quantities; for example, a quantity known 

 to be 0.03 magn. has been satisfactorily deter- 

 mined. One of the more important results was 

 to show that a photographic plate is more sensi- 

 tive when cold than when warm. At the same 

 time the scale becomes less, so that the character- 

 istics of a cold plate are changed to those of a 

 faster emulsion. This is of special importance 

 in winter, when plates may be taken from a warm 

 dark-room and exposed in zero weather. It is the 

 practise here to have all holders loaded in a cold 

 outer dark-room, or left outside long enough to 

 assume the external temperature before exposure. 

 The effect of humidity is to decrease the sensitive- 

 ness. The so-called "time correction" has been 

 investigated, and found to vary with the density 

 of the image, the character of the developer, and 

 other similar conditions. The light of the moon, 

 sun and several planets has been determined. The 

 light of the sky has been measured from noon to 

 ten o'clock in the evening. The difference in 

 brightness from day to night is about 17 magni- 

 tudes. The decrease at twilight is very rapid, 

 amounting to 10 magnitudes in an hour. The 

 measures of the bright stars have yielded perhaps 

 the most important result, the relation of photo- 

 graphic and photometric magnitudes to the class 

 of spectrum. The curve given in Vol. 59 of the 

 Harvard Annals was based on about 110 stars. 

 During the past year a redetermination of the 

 magnitudes of the 33 stars discussed in No. 4 of 

 Vol. 59 has been made. A supplementary list also 

 has been observed, which brings the number of 

 stars to 153. The latest results are included in 

 the figure exhibited. These values give the means 



of finding the photographic magnitude with con- 

 siderable accuracy, when the photometric magni- 

 tude and the class of spectrum are known. The 

 average deviation of the individual stars from the 

 curve is about ± 0.10 magn. 



The method of obtaining magnitudes by photo- 

 graphing stars out of focus has been safeguarded 

 so that there is little chance of grave error, except 

 as may be caused by a change of conditions occur- 

 ring during the period of exposure of any plate. 

 This is in part eliminated by requiring that five 

 measures of each star shall be made on five dif- 

 ferent nights. The distance at which the plate 

 is set from the focal plane precludes the possi- 

 bility of error arising from slight changes of 

 focus. Since the approximate photographic mag- 

 nitude is known in advance, it is possible to ob- 

 tain images that match closely those of Polaris 

 in density, thus making the scale of the plate of 

 less importance. The method thus far used makes 

 one exposure necessary for each star. Work on 

 the Pleiades has been begun, using a brass plate 

 having a number of apertures in it, so that each 

 star shines through its own individual window to 

 illumine the sensitive plate. Thus, a number of 

 stars may be photographed simultaneously. 

 Preparations have been made also to apply the 

 method of out-of-foous images to the region of the 

 pole. In order to obtain fainter stars the plate 

 will be very near to the focal plane, and settings 

 will be made inside and outside the focus in order 

 to correct for change of focus, lack of flatness of 

 the plate, and like sources of error. The great 

 advantage of all these methods is that they give 

 absolute values, and are free from many of the 

 errors that usually beset photographic work. 

 A Unique Perturbation of Neptune: W. H. PiCK- 



EBING. 



An examination of comet orbits led the author 

 to suspect the existence of a large dark unknown 

 body, located not far from the sun and in the 

 general direction of the north pole of the ecliptic. 

 Such a body should produce a peculiar perturba- 

 tion of the outer planets, forcing them to describe 

 smaller circles of the sphere, slightly to the south 

 of their assumed orbits. To determine if this 

 were the case a study has been made of the orbit 

 of Neptune, based on the observations made at 

 Paris and Greenwich. It was shown that from 

 1846 to 1873 the observations were well repre- 

 sented by Le Verrier's orbit and from 1882 to 

 1897 by that of Newcomb, but that Le Verrier's 

 orbit would not represent the later observations 

 nor the orbit of Newcomb the earlier, and sys- 



