March 12, 1896J 



NA TURE 



449 



"By an inspection of the numerous spectra my opinion is 

 strengthened that in arranging stars according to their spectra 

 only general marked characteristics should be considered, and a 

 rational classification can only l>e thought of if based on the 

 supposition that the various stellar spectra indicate different 

 phases of development. 



"I think it is much to be regretted that, in the extensive 

 spectroscopic examination of all stars down to about the seventh 

 magnitude, which Pickering by means of an objective prism has 

 undertaken, the classification of stars, based on no general 

 grounds, but only according to the appearance of the spectrum, 

 which is very often misrepresented by incorrect exposure, 

 especially in the case of brighter stars, leads him to adopt 

 sixteen classes, denoted by the letters A-Q." 



On this it may be remarked that Mr Lockyer has already 

 shovvn in Nati'RK that Prof. Pickering's classification is quite 

 philosophical, and that, moreover, it brings stars together to 

 which he also has assigned special places in his classification on 

 account of their special features. 



Prof. Vogel then continues : — 



" The attempt, on the lines mentioned above, which I made 

 twenty years ago, with regard to the classification of stellar 

 spectra has for the most part been corroborated, in spite of the 

 great progress of stellar spectroscopy during the last few years, 

 namely biy the fine detailed researches of their spectra by 

 Scheiner. 



"In relation to the stars of Class III., direct observation of the 

 less refrangible part of the spectrum of the photograph is yet 

 to be considered. Of my two suggested divisions, a and h, the 

 critcrium as to which belongs to the more advanced stage of 

 development is entirely absent One can only say this, that in 

 lx)th divisions the almospheres of the stars have so far cooled as 

 to stop the dissociation of the materials and allow compounds to 

 be formed. There is therefore, no reason given why stars of 

 Class III.(^, in which chiefly the absorption bands are produced 

 by carbon compounds, should be placed in a special Class IV. 

 For the same reason direct observation is a good means of 

 recognising spectra of Class II. Also here there is no reason to 

 adopt subdivisions other than the two already assumed by me 

 before we are in possession of more accurate investigations of 

 the spectra of Class W.b. 



" It is different in the case of the spectra of Class I. In these 

 spectra the application of photography makes it in general 

 possible to step further and obtain more minute points of 

 difference than was the case before. It appears also that the 

 study of the spectrum of these stars is of special interest, inas- 

 much as, starting from the simplest spectrum, in which only 

 hydrogen lines are visible, the first traces of a further develop- 

 ment can be found by the appearance of lines of other substances, 

 and we can follow them up to the numberless lines exhibited in 

 the spectra of Class II. Perhaps it will be possible by a more 

 ■ extended investigation of the details of the spectrum of Class I. 

 to find the first beginnings and individual parts of both from 

 successive series, the extremities of which, so different in 

 appearance, are spectra of the type of Class III. a and \\\.h. 



"The observations mentioned above have led me to the opinion 

 that the appearance of the lines of clevitc gas in stellar spectra, 

 if closely watched, may give us a good basis for the classification 

 of these spectra. 



"The spectrum appearance of clevite gas has such a 

 similarity with that of hydrogen, as has long been known to be 

 the case, through the constant appearance of the D^ line with 

 the hydrogen lines at all parts of the chromosphere of the sun, 

 as well as in the prominences, that one may expect to find, after 

 the appearance of the hydrocen lines in the first place, that of 

 the lines of clevite. The [jaucity in lines of the spectrum of 

 this gas makes it especially easy to be at once recognised. 

 Although the brightest line, A 3889 /z/i, as already mentioned, 

 fulls so near constantly the present hydrogen line 11^ in the spec- 

 trum of Class I. that a separation is impossible, and although in 

 only very few cases does the sum-total of both these strong lines 

 ajipear distinct— as, for instance, in the spectrum of & Lyra- — the 

 lines A.382"00/u/u, \386"8 ju/x, A 402'6juju, and A447*2/t/u, on the 

 ni her hand, and those in the less refrangible part of the spec- 

 rnn, namely A 492 2 ^/i, A 501 '6 /t^, and I),, A 587 '6 /u/*, are 



easy to find and recognise that the proof of the presence of 



vite gas is not be.set with difficulties. 



" As a second point of difference for subdivisions of Class I. 

 the presence of the calcium lines A.393"38/x/u and A 396 86 mm 's 

 convenient, the second of these lines coinciding nearly with 



NO. 1376, VOL. 53] 



the hydrogen line H« (A 397 '02 mm)- If the first of these lines be 

 thin and sharp, the second only influences the hydrogen line Wt 

 to a very small extent. If the lines of calcium increase both in 

 intensity and breadth, then lit broadens in a very distinct 

 manner ; both lines then very soon excel with respect to inten- 

 sity and thickness, the strongest and broadest hydrogen lines 

 in the spectra of Class I. On further development they form 

 together the characteristic pair of lines in Class II. which 

 Fraunhofer called H." 



Prof V'ogel then concludes his paper with a re-division of 

 stars of the first class into three parts, a, b, and c. 



Although (• includes stars having spectra which exhibit bright 

 lines, he suggests that from the present point of view it would be 

 perhaps better to place them first, as they represent the first 

 stage of development. The fact that they are here retained in 

 subdivision c is because a definite conclusion has not yet l)een 

 arrived at as to their true position, so the old position has, f-ro 

 tempore, been maintained. 



THE RONTGEN RAYS, 

 HTHE following is a summary of some of the work with 

 ■*■ Rtintgen rays brought under our notice during the past 

 week : — 



The results of a scientific study of the properties of Rontgen 

 rays are stated in a paper by Prof. C. Vicentini and Dr. G. Pacher, 

 in a paper read before the Reale Istituto Veneto di scienze, 

 lettere, ed arti, on January 26, and now issued as an excerpt from 

 the Meniorieni the Institute (vol. xxv. No. 7). The authors 

 found distinct evidence of an irregular reflection from a para- 

 bolic brass mirror ; the Crookes' tube and the sensitive plate 

 were placed on opposite sides of an iron plate, .so that the rays 

 to reach the plate had to be reflected from the mirror. No 

 effect was observed with a similar glass mirror, or with the 

 arrangements used by previous observers to test the existence 

 of reflection. Photographs accompany the paper, showing 

 shadows of fish, a hand, two feet, and a rat with mercury 

 introduced into its intestines. Experiments showed that a gold- 

 leaf electroscope, turned towards the kathode, became charged 

 positively ; a wire cage, however, completely annulled the 

 action. The following list oftransparencies to Rontgen rays is in- 

 teresting : — Solid bodies (opaque): Potassium, phosphorous, fused 

 sulphur, glass, sealing-wax, tin, zinc, iron, brass, copper, lead, 

 platinum, mercury, crystallised sulphur, rock-salt, quartz, calc- 

 spar, fluor-spar, topaz, beryl, aragonite, baryta, lead carbonate, 

 tourmaline, borax. Semi-transparent : Aluminium, sodium, 

 ebonite, retort -carbon, mica (along both axes). Transparent : 

 Cardboard, wax, paraffin, sugar, wood-charcoal, amber, shellac. 

 Liquid bodies (opaque) : Carbon disulphide, sulplyiric acid 

 (saturated solutions of), sulphates of zinc, copper, iron, cobalt, 

 nickel and magnesium, bichloride of mercury, chlorides of 

 sodium, ammonium and platinum, neutral oxalate of potash, 

 potassium bichromate, and ammonium nitrate. Semi-trans- 

 parent : Acetic acid, nitric acid, glycerine, ammonia, and, in a 

 less degree, distilled water and alcohol. Transparent : Ether, 

 benzene, in a less degree vaseline, petroleum, aniline, and olive 

 oil. There appeared to be no clear relation between the opacity 

 and density of the substances experimented upon. 



In order to get distinct shadows of objects opaque to Rontgen 

 rays, it is obviously necessary to avoid a penumbra, by employing 

 a source of the rays approaching as nearly as possible to a 

 point. The Crookes' tubes, however, as usually put on the 

 market, are given a pear-shape, for the purpose of having as 

 extended an area as jwssible under the action of the kathode 

 rays. To obtain clear shadows with such a tube the objects 

 must be very close to the sensitive plate or the fluorescent 

 screen, a condition which cannot always be realised if the ob- 

 jects are enclosed in envelopes of considerable thickness. Prof. E. 

 Salvioni described to the Accademia Medico-Chirurgio di Perugia, 

 on February 22, two methods whereby the kathode rays may be 

 brought to a point on the walls of the tube, producing a very 

 restricted but intensely fluorescent area. In Crookes" tubes, 

 with the usual spherical cup-.shaped kathode, the rays issue from 

 this normally, and meet in a point generally inside the tube. By 

 moving a magnet in the neighbourhood of the tube, this focus 

 maybe brought on to the tube wall. A more convenient methotl 

 of arriving at the same result was found by chance. On moving 

 the hand over the tube, the position of the focus of the kathode rays 

 is altered, and the .same action takes place on touching the tube 



