April, 1906 j 



KNOWLEDGE & SCIENTIFIC NEWS. 



405 



such modifications as the new arrangement requires. 

 The diaphragm has, of course, to be between the two 

 systems, and to maintain the aperture for oblique rays 

 the exterior lenses are shghtly enlarged and brought 

 nearer to the others. In such a complete objective the 

 light has to pass througli twelve surfaces of glass in 

 contact with air. And in a modification described, in 

 which there are eight separate lenses, there are sixteen 

 such surfaces. At every surface of glass to air there 

 is a loss of light due to reflection, and Mr. Taylor esti- 

 mates this at about forty-eight per cent, of the incident 

 light, when there are twelve glass surfaces. A part of 

 this light is altogether lost, passing out of the lens in 

 front, but a part goes in the other direction and con- 

 tributes to the general useless light in the camera, which 

 of course, tends to produce fog on the plate. Mr. 

 Taylor seeks to remedy this defect and secure greater 

 brilliancy of image, by tarnishing the polished surfaces 

 of the lenses, using a solution of sulphuretted hydrogen, 

 for example, until they assume a brownish-slate colour 

 by reflection. He thus claims to get a rather more 

 brilliant image with six or even eight lenses than with 

 only three lenses normally polished. It is quite easy to 

 understand how that such a procedure will reduce the 

 amount of reflected light, but it is not so easy to see 

 how it increases the image-forming transmitted light. 

 It appears at first sight that the tarnishing will reduce 

 both the reflected and the transmitted light because of 

 its absorbing power, but Mr. Dennis Taylor is such an 

 experienced and expert technical optician that, presum- 

 ably, there is some fault in this simple view of the 

 matter. It may be noted that these improvements in 

 Cooke lenses are not yet practically available. 



Clearing Intensified Negatives. — The character of the 

 gradation of a negative is, or ought to be, one of its 

 most valuable properties, and, as I have often said 

 befoi-c, there is only one method of intensification that 

 has been shown to preserve it, namely, the mercury and 

 ferrous oxalate method. All the more usual processes 

 and a great number of those suggested i>ut rarely used, 

 have been definitely shown to be uncertain and un- 

 proportional in their effects, and therefore to alter the 

 gradation in uncontrollable ways. A minor drawback 

 to the u.se of an oxalate when only hard water is avail- 

 able is the precipitation of calcium oxalate, some of 

 which may attach itself to the surface of the gelatine. 

 If kept clean, it disappears entirely on varnishing and is 

 quite inert, but it may, if desired, be removed by merely 

 placing the negative in weak hydrochloric acid until it 

 has been dissolved away. The strength of the acid 

 seems to matter very little; perhaps strong acid, with 

 ten times its bulk of water, is a good average. The 

 time required may vary up to a few minutes.' I have 

 seen celluloid films and plates of various kinds treated 

 in this way with invruiable success, and have not heard 

 of any failure or accident. The risk of frilling seems 

 to be practically negligible. Therefore there are three 

 distinct ways of obviating the difficulty resulting from 

 such a deposit :— (,.)_It may bo prevented bv the use 

 of soft water before and after the application of the 

 oxalate. (2.)— It may be removed bv means of hydro- 

 chloric acid. (3.)— It ni;iy be ncgkTfed if the negative 

 is to be varnished. 



E. A. Wilson. — We do not know of anv more recent book 



on tlin (Iftoriniiialii)n of longitude bv pholiigrapliic mp;uis 

 than thr oiu- vuu iiuntioii. 



ASTR.ONOMICAL. 



By Charles P. Butler, A.R.C.Sc. (Lend.), F.R.P.S. 



Determination of Radial Motions by 

 Objective Prisms. 



Director E. C. Pick!;ring has just issued a further circular 

 (Xo. 110) dealing with the method adopted at the Harvard 

 College Observatory for the determination of stellar motions 

 in the line of sight. The photographs are obtained with 

 the Draper Memorial telescopes used as prismatic cameras, 

 with large objective prisms placed outside the object glasses. 

 .\ photograph of all the stellar spectra included in the ree'ion 

 covered by one plate having been obtained, the objective 

 prism is then turned through 1800, and then a second ex- 

 posure on the stellar spectra is given on the same 'plate. It 

 is not necessary or advisable to use two plates, as was 

 formerly recommended. When, owing to special circum- 

 stances it is more convenient to reverse the telescope, instead 

 of turning the prism, the plate must in such cases be turned 

 iSoo._ The corresponding spectra of each star in the two 

 positions may be brought end to end, or in any desired 

 order, by adjustment with a cross-wired eye-piece'. If the 

 method were perfect it would only be necessar\- then to 

 measure the distance apart of the" corresponding lines of 

 each pair of spectra, and each star whose radial motion was 

 known would serve to determine the constant distance apart 

 of the lines ; the differences in distance, converted into wave- 

 lengths, would then give the required motion of the other 

 stars. .Since the motions of the sun and earth are the same 

 for all, these will be eliminated. 



The princii)al sources of error, such as those due to the 

 distortion of the lens, and temperature corrections, are 

 radial, and may be determined by using both co-ordinates 

 of the lines in all the spectra. .Any changes in the differ- 

 ential refraction may be reduced by turning the prism so 

 that the spectra become horizontal instead of vertical. 



To illustrate the method a sample plate of the Pleiades 

 group is given. On the scale of the spectra given bv the 

 1 i-inch Draper telescope the probable error in the determina- 

 tion of the motion would be + 3.5 km. In another series 

 obtained with the 8-inch Draper telescope, covering loo 

 square, the dispersion is about one-third of the former series, 

 but as the scale is also about one-third, the definition is 

 considerably better, and measures on these may have nearlv 

 the same degree of precision. Much fainter stars are shown 

 OTi the latter plates, lines being clearly defined in stars of 

 Ihu eighth magnitude. 



The Coming Total Solar Eclipse, 

 Ja-nuary 13 14. 1907. 



Of the six total eclipses which arc computed to occur 

 diiring the next six years, that of 1907 seems least uncer- 

 tain to yield signilic.ant results, and in anticipation of this 

 the local particulars for various st.ilions on the track of that 

 eclipse have bi'cn communicated to the Amvrkan Journal of 

 Sciciiir by D. Todd and R. H. Baker. 



Following the eclip.se of 1905, August 30, bv an interval of 

 seventeen months, the figure and "type of the corona will 

 doubtless have changed considerably, so that it is of the 

 highest imiHirtance to photograjih the solar surroundings at 

 this opportunity. Fortunately the track is wholly on land, 

 but a great part of the region is so remote and "difficult of 

 access, being in .Mongolia and the Gobi Desert, that it 

 could be occupied only by equipping tedious and expensive 

 < xpiditions. 



The western half of the eclipse track, however, traverses 



