June 15, 191 1] 



NATURE 



529 



reproduction of a composite drawing made from their 

 piiotograplis. On this drawing those parts which reflect 

 a greater amount of light from the red end of the spectrum 

 are shown in red, and those which reflect chiefly light of 

 shorter wave-length are represented in green. The com- 

 plex mixture of colour cannot be described in detail here, 

 but we note that the walls and several streaks about 

 Tycho are red, as are also the greater parts of the Mare 

 Serenitatis and the Mare Humorum. On the other hand, 

 Tranquilitatis, Fecunditatis, and Crisium are almost 

 wholly green, while the Mare Imbrium is quite an intricate 

 mixture. 



The Silver Disc Pvrhelio.meter. — In 1909 Mr. Abbot, 

 director of the astrophysical observatory of the Smith- 

 sonian Institution, designed a pyrheliometer which, 

 although not readily providing results reducible to calories 

 per sq. cm., furnishes readings proportional to the intensity 

 of the solar radiation, and can be made of a standard 

 form, so that a number of widely distributed observers 

 ma}- work in consonance. This instrument, as described 

 in No. 19, vol. Ivi., of the Smithsonian Miscellaneous 

 Collections, consists of a silver disc in which a radial hole 

 has been bored. Into this hole a thermometer bulb is 

 inserted and packed, a mercury filling ensuring good con- 

 duction. The whole is carried in a tube provided with 

 suitable diaphragms, exposing shutters, and packing, on 

 an equatorial stand, and the observations simply consist 

 of reading, very carefully, the thermometer at definite 

 intervals. In order to secure cooperation, the Smithsonian 

 Institution has furnished several other institutions with 

 copies of the standard instrument, and is willing to 

 furnish others. Standardised and packed ready for ship- 

 ment, such an instrument costs the institution about 100 

 dollars. 



The Photographic Determination of Stell.^r Magni- 

 tudes. — 'Two sections of the Harvard College Observatory 

 Annals recently received deal with the difficult Lut 

 important question of the determination of stellar magni- 

 tudes by photography. 



In No. I, vol. Ixxi., Prof. E. C. Pickering discusses the 

 numerous difficulties attending the determination of 

 standard magnitudes, and briefly reviews the various 

 methods that have been tried. He then shows by tables 

 and curves that the method adopted at Harvard, in which 

 the photographic magnitude can be derived from the 

 photometric by adding a constant depending upon the 

 class of spectrum of the star, gives results in which the 

 deviations, at least for the brighter stars, are probably too 

 small for determination at present. A comparison of the 

 results obtained by thus treating the magnitudes given in 

 the Uranometria O.xoniensis and the Potsdam Catalogue 

 shows the smallness of the residuals. The photographic 

 magnitudes of all stars of mag. 4-25 and brighter from 

 these two catalogues and the Harvard Revised Catalogue 

 are given in Table V., and should prove a useful refer- 

 ence for standard photographic magnitudes ; other tables 

 give the magnitudes of double stars and of stars having a 

 photographic magnitude of 3-0 and brighter. The paper 

 concludes with an interesting note on the apparent changes 

 well-known groups of stars would undergo if we saw their 

 photographic, instead of their photometric, brightness; 

 Betelgeusc would become from o-;; to i-o mag. fainter 

 than the belt stars ; .Xntares would appear a little less 

 bright than a and t Scorpionis, and the Southern Cross 

 would be greatly altered by 7 Crucis becoming much 

 fainter than 5. 



In \n. G, vol. lix., Mr. E. S. King gives the photo- 

 graphic magnitudes of 153 stars determined by the method 

 of measuring images taken at different foci : these magni- 

 tudes are absolute, and the scale is independent of any 

 other svstem. The determination of these absolute magni- 

 tudes provides a surer basis for the derivation of the 

 constant corrections to be applied to existing photometric 

 magnitudes according to the class of spectrum, and these 

 corrections are tabulated. Mr. King also discusses his 

 material to educe evidence of the absorption of light in 

 snace, and finds, from the increased redness of the more 

 distant stars, that there probably is some such absorp'-'^'i 



Jupiter's Eighth Satf.i.ute. — The eighth satelT 

 Jupiter was observed at Heidelberg on May 22, ,1;,.' ... 



NO. 2172, VOL. 86] 



f*^ 



loh. 26-2m. (Kdnigstuhl M.T.) its position (1855-0) was 

 i4h. 19m. 37-4S., -13° 56-8'. Its magnitude was between 

 16 and 17, and Dr. Wolf remarks that its position is so 

 low as to place its observation with the reflector on the 

 limit of possibility, mechanicallv i^A.stronom\sc]\e Nach- 

 richtn,, \o. 4503). 



THE BRITISH SOLAR ECLIPSE EXPEDITION.' 

 H.M.S. " Encounter." 



April 29, 

 Eclipse Day, 11.30 p.m. 

 'T'HE eclipse is over, and we have only seen quite a little 

 of it. For a long time we have been making our 

 preparations, and for more than a week we have been drill- 

 ing nearly daily. The weather has not been very kind, and 

 either heavy showers or light rain have fallen during the 

 day or night. However, it seemed tli.it \vc might be lucky 

 and secure a cloudless moment, or we might repeat the 

 Palma eclipse of 1905 and be partially clouded out. Well, 

 it has been very much worse than Palma. 



The early morning of the eventful day was far more 

 cloudy than usual , in fact, there was a "great amount of 

 cirrus cloud and numerous cumuli. As eclipse time came 

 along, evidently the reduction of the temperature in this 

 very humid climate (nearly always 75 per cent, or more of 

 saturation) caused the aqueous vapour to Condense into 

 cloud, and the cloud increased and increased in amount 

 as the time for second contact approached. So cloudv was 

 it, and so disturbed was the air, that the observer who 

 was watching the cusps could not be certain of giving the 

 correct observed time of totality by them. 



By chronometer time the eclipse clock was started, and 

 the whole party went through 

 their programme as if it 

 ■vere a rehearsal. The 

 greater part of totality was 

 hardly observed at all, and 

 the whole of it was seen 

 through fairly thick cirrus 

 cloud. After the eclipse was 

 over the clouds gradually 



dispersed, and long before // • 



fourth contact was reached a M9>r*-yy\ 



nearly cloudless sky was en 



evidence. The whole affair has been a great disappoint- 

 ment, both to the astronomers and the ship's company. 



The corona was distinctly of a minimum type, the fish- 

 tail portion ol ii being pointed towards the zenith (see 

 sketch). Om l.irgc prominence in the north-west qu.idrant 

 was conspicuous towards the end of totalit\, and gave a 

 ruddy hue to the corona in that quarter. Owing to the great 

 amount of cloud, long extensions could not be observed, 

 but the fish-tail portion extended, to my eye, one diameter 

 away. Clouds prevented me from seeing any conspicuous 

 feature on either the east or west limbs, but on the lower 

 limb there seemed to be the indication of the stump of a 

 large streamer. 



.At the time of writing, the following results have been 

 secured, but only a few plates of some of the instruments 

 have been developed. Steward equatorial, three plates 

 developed, two of which may be useful. These show the 

 lower corona only, including the stumps of some streamers. 

 Two platps only of the De la Rue coronograph have been 

 developed, and they are not dense, owing to the cloudy 

 nature of the sky. One plate from the 4-inch 16-foot 

 coronograph has been developed, and this seems very much 

 underexposed. No plates of the large grating spectro- 

 graph have as yet been developed, but one (the only plate 

 developed), taken with the 6-inch prismatic camera near 

 the end of totality, shows a considerable nimiber of lines in 

 addition to some continuous spectrum. In fact, the photo- 

 graphic results so far obtained seem to be better than what 

 was anticipated from the cloudy conditions. 



One may really sum up the whole results of the expedi- 

 tion as a failure, considering what we might )'.i\ 

 obtained had the weather been more propitious. 



Nearly all. if not all, the eclipse parties have obtained 

 almost simil.'ir results ; but it is early yet to say for certain 

 nntil the pl.it">; haw been developed. 



W T. '^. r.ocKVi:R. 



' Continued from p. 4y,.,. 



