January, 1907.] 



KNOWLEDGE & SCIENTIFIC NEWS. 



13 



Talbot's Law. 



because the film has become more dry or set, or the 

 light may cause a chemical change in the material and so 

 tend to retard the further migration of the compound. 

 To confirm the results, a compound of silver was used, 

 applying it to gelatine films which were then shielded in 

 alternate strips and exposed. The silver was estimated 

 in the exposed and unexposed strips, and the former 

 contained sometimes twice as much as the latter. 



The moving about of a substance that is free to move 

 is very familiar in crystallisation; the spaces between 

 the crystals being ne;irly, if not quite, free from the sub- 

 stance, while at first it was evenly distributed through 

 the solution or space it occupied. It has often been 

 observed that light appears to facilitate or encourage 

 the deposition of a solid, though it has been stated, 

 without, perhaps, sufficient evidence, that the effect is 

 due to the shaded part being of slightly higher tempera- 

 ture than the other. The experiments referred to above 

 appear to support the idea that light itself is really 

 effective. 



The rotating disc with differently coloured 

 sectors is so often used in colour photo- 

 metry for photographic purposes, that the accuracy of 

 this method of work is a matter of prime importance. 

 So far as visual work in concerned, it depends upon 

 Talbot's law, which states that " if any part of the 

 retina be excited by intermittent light recurring regu- 

 larly in the same way, then, if the period be sufficiently 

 short, a continuous impression will result, which is the 

 same as that which would be produced if the total 

 light were distributed uniformly throughout the whole 

 period." It is satisfactory to know, that investigations 

 by Mr. E. P. Hyde, which he described before the 

 American Physical Society last .April, show that this 

 law is correct for white, red, green, and blue light for 

 angular openings from 10° to 288°, within a possible 

 error of .3 per cent., though the deviations from the law 

 may be a little greater in the case of the coloured lights 

 than white light. Therefore the doubts that have from 

 time to time been cast upon the accuracy of this method 

 appear to be without foundation. 



I use the word '' still " here as the 

 1^*'" t anthithesis of " rocked," the developing 

 ■ solution not being moved over the sur- 

 face of the plate. This method of development 

 has recently been described and eulogised as 

 if it were new, but it is really quite old and the effect 

 of it has long been well known. It may be car- 

 ried out in more ways than one. The plate may be 

 allowed to lie at the bottom of the developer, which 

 is preferably made rather weaker than usual, though 

 not too weak, or the mixing of the various parts of the 

 solution by diffusion will defeat the very object of the 

 method; or as soon as the plate has absorbed a suitable 

 quantity of the developer it may be removed from the 

 dish so that the only developer that is available is either 

 in the film or on its surface. It is clear that by such 

 treatment the developer available on the most exposed 

 parts will be more quickly exhausted than that on the 

 less exposed parts, and that therefore the growth of 

 the image will slacken most where the exposure effect 

 has been greatest, and the result will be a flatter (or, 

 to use a more complimentary term, a " more har- 

 monious ") negative. AV'here there is a tendency to 

 hardness, as from under exposure, this method may be 

 useful, but generally it is far better to keep the de- 

 veloper moving so that it may have an equal energy 

 over the whole plate, and the extent of its action may 

 be governed only by the exposure effect. 



ASTRONOMICAL. 



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



Photog:raphic Photometry of Coloured 

 Stars. 



Mr. R. J. Wallace, at the Verkes Observatory, has been 

 making further investigations into the suitability of em- 

 ploying colour screens for stellar photography. In the pre- 

 sent paper his remarks are more particularly directed to 

 photography with a reflector, so that the chromatic aberra- 

 tion of the instrument itself may be neglected. It is well 

 known that with almost all kinds of photographic plates 

 there are certain portions of the spectrum to which the plates 

 are less sensitive than others. So long as the star groups 

 being photographed are all of the same type of spectrum, 

 then any kind of plate will give a true rendering of their 

 relative brightnesses. In reality, we know that in any star 

 field there are many types of stars, and to obtain a true 

 photographic record of their magnitude we must adapt a 

 colour screen in front of the photographic plate which will 

 reduce the intensity of that kind of transmitted light to 

 which the plate is specially sensitive, so that it may record 

 the true colour values of the object. The dyes Tartrazine, 

 .\uramine O, Metanil Yellow S, and Xitrosodimethylanilin 

 were found to be most suitable, and the requisite amounts 

 were incorporated in gelatine solutions and flowed on pre- 

 pared glass plates. In the case of compound filters, it was 

 found preferable to coat several plates with only one colour 

 each, and then to bind the plates together. The perform- 

 ance of the filter is accurately examined by means of a 

 delicate spectrophotometer, and alterations made until it is 

 exactly adapted to the photographic plate to be employed. 

 Reproductions are shown of photographs of the region sur- 

 rounding the coloured star U Cygni (which is classified by 

 Chandler in his colour scale as 9.3, where 0= white), first on 

 ordinary Seed 27 plates, and then on Cramer isochromatic 

 provided with the proper colour filter. The difference is 

 very great, and emphasises the danger of determining any 

 star magnitudes from photographic records unless there has 

 been such control over the colour selection of the plate. — 

 (Astrophysical Journal 24, November, igoli.) 



Tnunderstorms and Sunspot Frequency. 



.•\ considerable amount of discussion has been published 

 in the endeavour to prove the existence of direct correspond- 

 ence between the occurrence of large thunderstorms and 

 the state of the solar activity as evidenced by the epoch of 

 sunspot frequency. Quite recently Dr. A. S. Steen 

 examined the data from a large number of stations in 

 Norway, Sweden, and Denmark obtained during the period 

 1873-1903. He concludes that the epochs of maxima and 

 minima of thunderstorms occur at or near the epochs of 

 maxima or minima of sunspot frequency. In addition to 

 the main period of eleven years, a subsidiary period of 

 about 5.5 years is also indicated. — {Hanii-Band dcr Mcfcmo- 

 lof]ischeii Zdtschrift.) 



Dealing with the same subject, Mr. .\. Mutt announced 

 at a meeting of the British .-Xstronomical .Vssociation that 

 on almost every occasion in his experience on which there 

 had been a thunderstorm, a new spot had broken out on 

 the sun on that or the next day, or that one of the already 

 existing spots h.ad greatly increased in magnitude. .\n in- 



stance of this occurred on Mav 



1906. — (Journal Brit. 



Ast. Assoc. 17, November 22, uioCt.) 



Echo of the Valparaiso Earthquake. 



The seismograph records ol llie ^l.ltions along ilu- Pacific 

 Coast of the Asiatic Continent are of special interest when 

 they occur in connection with the convulsions along the 



