Januabv, ia02.] 



KNOWLEDGE 



11 



to li' in seven weeks, or 11' annual proper uiotiont+ ; 

 and this at a distance uo greater than that of the 

 nearest fixed star means a velocity of 2090 miles in the 

 second. 



But difficulties begin with the persisting width of 

 the bright hydrogen lines. Unlike their dark com- 

 panions, these remained strong and broad to the end 

 of April, and again from August 24 to November 28 

 they were, still, as broad as they were in the middle 

 of April, but uo longer the strong lines of the spectrum. 

 For nine full months they have remained abnormally 

 broad; and, since April 16, they have retained constant 

 widths, as compact bands without their hazy edges. 



A line may bo widened by pressiu-e of the radiating 

 gas, or by the two-way displacements effected by rotating 

 velocities about an axis at right angles to our sight-line. 

 It is not easy to conceive the physical conditions of a 

 star, in which the whirling velocities of the heated 

 hydrogen atmosphere could be maintained practically 

 constant for nine mouths. Pressure, on the other hand, 

 only requires the central mass to be gieat enough ; but 

 the greatness is too great to be easily accepted; and the 

 compact appearance of the lines is not an expected 

 result of pressure, which hardly admits of uniforai 

 density with sharp edges. The dilBculty is great; but 

 it does not threaten one more than another hypothesis 

 coacerning the origin of the new star in Perseus. It 

 seems rather to be a sign of new discoveries in the 

 spectrum of hydrogen to be expected before any safe 

 conclusion can be drawn about the physical condition 

 of the star-; and other peculiarities of the fourth 

 spectrum point the same way. 



The fourth spectrum was not photographed at Stony- 

 hurst until August 24. 



All the lines of this spectrum agree, in positions, with 

 those obsei-ved in gaseous ncbulse. But they are all too 

 broad, and four of them, 3869, 3969, 4363, and 4719, 

 are composite bands, the first three of four components, 

 and the fourth of three. 



But apart from the question of the nebulous character 

 of the star, the peculiarities of this spectrum are to us 

 at present an enigma, of which the key-word is probably 

 the lino 3969. This line began to put on its present 

 disguise at a rather later stage of cooling than its com- 

 panion line 3869, at a temperatui-e coiTcsponding with 

 the magnitude 5.3, by appearing in greater strength 

 than the hydro<:ren line H8 on April 16. If it is the 

 hydrof,'en line Hs, what condition of i,'lowiu<ir hydrogen 

 is indicated by this line being now stronger than all the 

 rest taken together, when usually it is the weakest of 

 those present in the photogi-aphs? It might be the 

 calcium line known as Fraunhofcr's H in the Solar 

 spectrum ; and if so, under what conditions can the 

 calcium H glow strongly without its inseparable and 

 stronger companion K? It might be the helium line 

 3965 greatly widened towards the red side. Three 

 helium lines could then be counted in the spectrum 3S08, 

 3965, and 4713, the two latter similarly widened so 

 much to the red side as to show the apparent centres 

 of the bands at 3969 and 4719; and all three are com- 

 posite bauds. But why ai-e these the selected lines, or 

 why are the stronger ones absent? 



The answers to these queries, when found, will define 

 securely the physical condition of the star, at a time 

 when its spectrum looks so like and yet so unlike that 

 of a gaseous nebula. 



tt Lick Obs. BuUetiu, No. lo. 



Hettrrs. 



[The Editors do not bold themselves responsible for the opinions 

 or statements of correspondents] 



NOVA PERSEI. 



TO THE EDITORS OF KNOWLEDGE. 



Sirs, — The interesting note by M. Antoniadi on the 

 photogi-aphic image of Nova Persei is quite in accord 

 with exposures I have made with a portrait lens by Ross. 

 I find the second aiu-eola, which is shown when the 

 exposure has lasted 3A hours, seems to be about 2i times 

 further off the central dot than the first aureola, tho 

 latter being well shown with an exposure of only ten 

 minutes. If a part of the object-glass is covered tho 

 portion of tho imago on the same side of the disc is 

 missing. This effect can also be seen in other stai-s, 

 especially those v^fhose photographic image is larger than 

 would be expected from their visual brightness, like 

 S Persei. If, now, the plate is withdrawn outside the 

 main focus, say, for about If in., the images of most 

 stai-3 will be neai-ly doubled in size, and will have a 

 central dot surrounded by a penumbra equally darkened 

 to the outer edge, but the image of tho Nova will be 

 nearly of the same size as at the principal focus of 

 the lens, and will be made up of a central dot with a 

 ring round it and the space between only lightly dis- 

 coloured. This effect is more visible perhaps when pait 

 of the aperture is covered, when the central dot and 

 part of the ring do not appear connected. Tho plato 

 ijeing so far behind tho focal plane it will be seen that 

 the part of the image obliterat-ed by partly covering the 

 object-glass is the opposite side, not the same one as in 

 the chief focus. It seems as if there should bo two 

 bands in the spectrum beyond the part which causes tho 

 central dot, the second one being two or three times 

 further off and of a weaker photographic effect. The 

 spectra shown at the last meeting of the R. A. S., if I 

 remember right, consisted of the nebulous bands and 

 two bright thick lines high up in the violet, of which 

 the second or further one was the brightest, and, more- 

 over, both bands were nearer together than to the main 

 part of the spectrum, which does not seem to correspond 

 to the size of image obtained with a portrait lens. 

 Inglefield, Little Heath, Potters Bar. Henry Ellis. 

 24th November, 1901. 



RAINBOW BEFORE SUNRISE. 



TO THE EDITORS OF KNOWLEDGE. 



Sirs, — On the moniing of Thursday, 21st November, 

 I noticed gorgeous colouring in the south-east part of 

 the heavens, prepai'atory to sunrise. Near the horizon 

 the sky was comparatively clear, but higher up was a 

 lai-ge patch of cloud, tinted brilliantly with red and 

 orange, the intervening spaces being blue with shades 

 of green. There was at the same time a shower of rain 

 advancing from the west, which caught the bright light 

 reflected from the cloud, with the result that a piu'plo 

 rainbow became visible. The time of sunrise for this 

 day is 7.29, and the rainbow was visible from 7.15 (or 

 possibly sooner) to 7.30, when it began to fade away ; 

 before 8 o'clock rain was falling overhead. The bow 

 was of very large dimensions, with its two extremities 

 at right angles to the horizon. I should be very much 

 interested to know whether any of your readers have 

 a recor-d of a similar rainbow before sunrise. 



E. W. Johnson. 



50, Birdhurst Road, South Croydon, 

 30th November, 1901. 



