July 27, 191 1] 



NATURE 



Single with K 3 , it seems to me risky to base certain 

 reasonings on the appearance of the resulting spectrohelio- 

 grams. Accepting Prof. Hale's interpretation, given now 

 iii.inx ,'..,- ago, that K, represents the highest le\ 1 and 

 K. an intermediate one, I submit that the Meudon plates 

 are more likely to represent the true spectroheliographic 

 aspect of the sun. It may be, as Mr. Evershed says, that 

 the dark concentrations called fiocculi are entirely due to 

 variations in the intensity of the narrow absorption line- 

 ar, in my opinion, rather the other way about) ; but is 

 not this variation due, it not entirely, at least to a great 

 extent, to the presence or absence or degree of intensity 

 of the K a radiations on either side of K 3 ? A study of 

 Al. Deslandres' spectroheliograms taken on the sectional 

 principle leads one irresistibly to think so. From the pre- 

 Eding remarks anyone can gather why I considered, and 

 still do, the assumption of alternating appearance and 

 Bsappearance oi the large flocculus covering the range of 

 ces during March and April, iqio, rather unlikely. 



As regards the points raised by me about absorptively 

 acting clouds seemingly cutting off the range of promin- 

 ences sharply at one common level, I must adhere to my 

 statements. They are the result of repeated direct 

 observation, and the phenomenon was strikingly on view 

 again only as recently as April 26, when a fine range 

 of prominences on the north-east limb showed it fairly 

 well. I have given in The Observatory recently a sum- 

 mary of my observational experiences up to date, and 

 amongst other matters also refer to the often seen pheno- 

 menon of dark matter being interposed between solar 

 prominences and the observer at levels attained by the 

 luminous portions of the prominences themselves. 'Such 

 observations have by no means remained unique, and I 

 'possess, through the kindness of Mr. Slocum, a pair of 

 excellent photographs depicting it in a case of some fine 

 prominences seen during October, 1910. As regards my 

 several visual observations of apparently overlying flat and 

 dark clouds (darker than the general dark tint of the 

 surrounding sky) abruptly stopping the bright upwards 

 flowing prominence vapours, I feel certain that if Mr. 

 Evershed had been at my side at the time his practised 

 eye would not have failed him to see what I saw, but 

 whether he would have been able to record it on a photo- 

 graphic plate in monochromatic light (Ha) without an 

 fidipse I hesitate to say. I hope, however, that the hypo- 

 thetical overlving cloud will not be taken as of the nature 

 of " smoke " over a fire. In my opinion it is more a 

 case of the rather abrupt entry of rising hot vapours 

 into a well-defined, more or less horizontal, stratum of 

 considerably less temperature, and that at a comparatively 

 abnormally low solar level. I readily admit that the cases 

 are few where the circumstances necessary for the pheno- 

 menon favourably combine with the all too rare cases of 

 the equally necessary perfect definition. 



Albert Alfred Buss. 



" Barrowdale," 22 Egerton Road, Chorlton- 

 cum-Har'dy, Manchester, May 13. 



If it is true, as Mr. Buss suggests, that variations in 

 the intensity of the components of the emission line K, 

 on either side of the absorption line K, are sufficient to 

 account for the dark markings occasionally found in 

 spectroheliograms, then the Kodaikanal plates should 

 show them as conspicuouslv as those taken with a hie;h 

 dispersion instrument, which isolates the central line. Vet 

 this, as Mr. Buss has himself pointed out, is not the case. 

 In studying high dispersion spectrum photographs of the 

 . iTie occasionally discovers places where the K, 

 line i- abnormally dark, and the same thing may also be 

 well observed in the line H«. When the spectroscope slit 

 chances to cross one of these linear markings, an intenselv 

 Rack spot is seen on the absorption line, and this will 

 usually remain visible or run along the line if the solar 

 imaev is moved slightly. In the case of the lines H and 

 K. the components of the emission lines H, and K, are. 1 

 think, always weak at the points' of greatest darkness in 

 the absorption lines, and for this reason they may possibly 

 contribute somewhat to the final result in our plates. 



Tli' intermittent character of the absorption marking 

 gescribed by me in The Astrophysical Journal for January 

 is, I think, demonstrable from a careful study of our 



spectroheliogram negatives, notwithstanding the fact that 

 these plates are of a somewhat composite character, repre- 

 senting the sun in K, and K, radiations. The disappear- 

 ance of the enormously extended marking between March 

 25 and 2b, 1910, could be accounted for, it is true, on the 

 supposition that in the interval between these days there 

 was a development of velocity in the line of sight exceed- 

 ing 15 kilometres per second; this would alter the wave- 

 length sufficiently to throw the dark K 3 line entirely off 

 the camera-slit. But such motion in a prominence usually, 

 if not always, presages a complete dissolution. 



Instances of the rapid disappearance of tie-, L urious 

 absorption markings are not infrequently met with in Ha 

 spectroheliograms, which show them so much more clearly 

 than do the low dispersion calcium plates. Since com- 

 pleting the construction of the new auto-collimating spectro- 

 heliograph of this observatory, 1 have obtained a nearly 

 continuous daily series of Ha plates during April and 

 May of this year. These are taken with the camera-slit 

 adjusted on the central portion ol the line, and represent 

 the highest levels on the sun. The images show most of 

 I the prominences as absorption markings on the disc, and 

 some of them are so dark as to appear like clear glass 

 in the negatives. Already in this short series several 

 cases have been noted of the disappearance within twenty- 

 four hours of very large masses of absorbing material. 



An interesting example was photographed on May 27 at 

 2h. 28m. and 2h. 53m. a.m. G.C.T. The disc of the sun 

 in these plates appears to have had a large letter S 

 engraved upon it with great distinctness. If drawn out 

 into a straight line, the S would measure more than 

 150,000 miles in length. On the following morning we 

 examined the plates with great curiosity to see what the 

 next letter might be ! The main portion of the marking 

 had, however, entirely disappeared, and only a few small 

 patches remained. This marking came into being with 

 equal suddenness, for an excellent plate taken on May 26 

 shows no trace of it. 



While agreeing with Mr. Buss as to the occasional 

 presence of small patches of absorbing matter interposed 

 between a prominence at the limb and the observer, I 

 am sorry that both our visual and photographic records 

 are entirely at variance with him with regard to the sup- 

 posed absorbing clouds overlying certain prominences, 

 which appeared to Mr. Buss to be cut off at one common 

 level. I have before me the K-line negatives and the 

 drawings in Ha of the prominences of March 17 and 18, 

 iqio, and April 26 and 27, 1911, both of which Mr. Buss 

 has cited as instances. These prominences were observed 

 here and photographed under almost as good conditions 

 as can be had at 7700 feet altitude, and the photographs 

 show a mass of detail in the higher parts, especially in 

 the prominence of iqio. Vet there is no trace of any 

 such appearance as Mr. Buss has described ; the highest 

 filaments rise to many different altitudes, both on the 

 drawings and photographs. 



I may perhaps mention that reproductions of our photo- 

 graphs of the iqio prominence, as well as some of the 

 solar disc showing the absorption markings, have been 

 sent as an exhibit to the Indian Section of the Festival of 

 Empire. J- Evershed. 



Kodaikanal Observatory. June 12. 



Hamilton and Tait. 



Though I did not miss the passage in his Life of Tait 

 10 which Dr. Knott refers in Nature of July 20 ("p. 77I. I 

 forgot about it when I wrote my review. The point as to 

 Hamilton's activity in quaternionic work is not of very 

 great importance, but my statement is borne out by 

 Graves's Life of Hamilton, which I read long ago, and 

 lin referred to, as well as by the published corre- 



d rice. Tait's introduction to Hamilton took place in 



[858; Craves states (vol. iii., p. 07) that Hamilton allowed 

 himself to be diverted in rSj7 from quaternions — the task, 

 , .if writing the " Elements " — by the subject of 

 definite integrals. Vccording to Dr. Knott, Hamilton^ did 

 not begin the composition of the "Elements" until a 

 eood deal later, and this view would appear from Dr. 

 Knott's statement, and from Hamilton's own language in 

 liis lr iters, to be correct. A. G. 



NO. 2178, VOL. 87 



