Dec. i6, 1875I 



NATURE 



129 



The + and - below show the mean error in that digit to exceed 

 or fall short of the mean of all ( I "44). The readings are on a I -inch 

 space. Thus it is apparent that the errors exceed the mean in 

 the four middle digits, and fall short of it in the others with but 

 one exception. 



Now, the practical results which may be deduced from these 

 data are as follows : — 



From (i), that there is considerable variation in accuracy of 

 estimation ; yet, in persons fit to estimate, it may be put at 

 </ -f- 50 up to ^ 4- 100 ; and therefore as the maximum error in 

 some himdreds of readings is only about three times the average 

 error, it is useless (and, worse than that, confusing) to have 

 di%isions smaller than 20 x the quantity required to be read. 

 The application of this result to surveyors' rods, &c., would be 

 very conducive to ease in distant reading. 



From (2), that for ordinary eyes about / -^ 2000 b the mean 

 error of reading (with one eye only), and therefore that i i\%^ 

 is the smallest size useful for reading ^sith the naked eye ; conse- 

 quently any closer divisions than these (^ of an inch) are detri- 

 mental because of the greater confusion they introduce. 



From (4), that in small divisions no consideration is neces- 

 sary, as the first instinctive impression is nearly, or quite, as 

 accurate as careful estimation. 



From (5), that it is important not to look at a space askew ; 

 and that if imavoidable, careful consideration is desirable. 



From (6), that in reading a series of observations of one 

 amount, the right and left eyes should be used alternately to 

 equalise the + and - errors. 



From (8), that the right eye should be used in preference to 

 the left. 



From (9), that both eyes should be used wherever possible. 



From (10), that red is the worst colour for measures, and that 

 green or blue are as good as white ; so that blue and white 

 would probably be the best practically, and more distinct firom 

 vegetation, &c., than the green at a distance. 



From (11), that more care is required in readings in the central 

 portions of a space than at the ends. This is as true on spaces 

 of / -f- 41 as on 10 J spaces. 



Of course some of these results may be personal errors, and 

 not generally true, but at least this notice will perhaps call the 

 attention of those accustomed to estimation to the subject, and 

 so elicit some definite statistics. There are many other interest- 

 ing questions which I have not touched on here as being of less 

 consequence, and not affecting practical questions so much as 

 the foregoing results ; these results it is to be hoped may lead to 

 a better acquaintance with a subject of such practical im- 

 portance. W. M, Flinders Petrie 



Bromley, Kent 



The late Dr. von Willemoes-Suhm 

 Ix the list of papers given in Prof. Wyville Thomson's 

 obituary of the late Dr. von Willemoes- Suhm (Nature, vol. 

 xiii., p. 88), is omitted — " Notes on some young stages of Um- 

 bellularia, and on its Geographical Distribution ;" (" Ann. and 

 Mag. Nat Hist.," 4th ser., vol. xv., p. 312, and pi. xviii. A.) 

 This animal is a Pennatnlid, and was obtained during the Chal- 

 hnger's cruise at depths from 1,375 to 2,600 fathoms. It is 

 usually associated with decidely deep-sea forms. 



To this may be added two papers which were read at the meet- 

 ing of the Royal Society on Dec. 9th, " On the Development of 

 Lepas Fascicularis, and the MrtA/lwrtj' of Cirripedia ;" and 

 " Preliminary Remarks on the Development of some Pelagic 

 Decapoda." J. C. G. 



Seasonal Order of Colour in Flowers 



I HAVE often noticed during spring and summer that flowers 

 appear to follow the spectrum firom the blue of the wild li3racinth 

 to the deep scarlet of our summer flowers. Will any of your 

 correspondents tell me if during spring the actinic rays of the 

 sun are in greater force, and whether these chemical rays are 

 those which revivify seeds and plants after their winter's hiber- 

 nation ? During autumn I have noticed the same gradations in 

 the colours of flowers. C. E. Heron Rogers 



Retford, Notts, Dec. 7 



OUR ASTRONOMICAL COLUMN 

 Atlas Pleiadum, 2 453. — Has anyone provided with 

 a large telescope examined this object of late for dupli- 

 city ? It appears very improbable that Struve's epoch of 



1827 can have been put upon record through an illusion. 

 We find in " Mensurse Micrometricae,*' p. 2 : — 



1827-16 Position i07°-8 Distance o''79 Mag. 5 and 8. 

 It is then termed " duplex difficillima." On subsequent 

 examination, i830'25, the position was fovmd to be 29*2, 

 and the distance only o"'3S. In 1831 and 1832 it was not 

 seen double with the Dorpat refractor imder a power of 

 600. Struve remarks, "si justa est anni 1827 notatio, 

 quod vix dubitare potest, cum nullum observans dubium 

 expresserim, motus in hoc systemate egregrie celer inesse 

 debet" From 1840 to 1847, Madler observing with the 

 same instrument, says he always found the star " quite 

 round," and to our knowledge a few years since the star 

 was pronounced single after close scrutiny with more 

 than one of the larger telescopes. 



The " Gegexschein " of the Zodiacal Light. — 

 The last number of the " Monthly Notices of the Royal 

 Astronomical Society " contains some interesting remarks 

 by Mr. Backhouse, of Sunderland, bearing upon this 

 phenomenon of the zodiacal light, as observed by him 

 between the years 1871 and 1875, which are the more 

 notable, as the observer does not appear to have been 

 aware of its previous observ^ation, and consequently the 

 particulars he has placed upon record afford a very satis- 

 factory confirmation of the existence of this feature of the 

 light. The " gegenschein " consists in an elliptical patch 

 of light almost directly opposite the sun's place, with the 

 greater axis nearly in the plane of the echptic. Attention 

 was first directed to it in the year 1854 by Herr Brorsen 

 (the discoverer of the comet of short-period — February 

 1846) — while obser\-ing at Senftenberg in Bohemia. It 

 may be inferred that it is generally difficult to distinguish 

 or trace, with certainty, as even Prof. Heis, of Miinster, 

 with his excellent vision and long practice, did not suc- 

 ceed in verifying Brorsen's obsen'ations for several years 

 (Zodiacallicht-Beobachtungen, Heis, 1875). Observations 

 of the phenomenon will be found in various numbers of 

 the " Astronomische Nachrichten." Brorsen thought this 

 opposition-glow of the zodiacal light was more distinctly 

 seen near the vernal than the autumnal equinox ; the ob- 

 servations of Heis are in January, April, and in December. 

 There are also observations by Schmidt at Athens, Schia- 

 parelli at Milan, &c. 



The Comet of 1729.— This very remarkable body 

 presented a case which is quite unique in the historj- of 

 these interesting denizens of the solar system. It was 

 visible without telescopic aid, or with little optical assist- 

 ance, at distances from the earth and sim not greatly 

 inferior to those to which any other comet has been fol- 

 lowed even with such an instrument as the Poulkova 

 refractor, and it may be inferred was of very different 

 constitution. 



The comet was first detected by Sarabat, an eccle- 

 siastic of the Jesuit order, at Nismes, on the 31st of July ; 

 the moonlight, which was then on the increase, prevented 

 his seeing it on the following evenings imtil Aug. 8, when 

 the moon being totally eclipsed, he recovered the comet 

 and gave intimation of his discovery to Cassini at Paris, 

 by whom it was observed on fort>--four nights between 

 Aug. 31 and Jan. 18 following, Cassini saw the comet 

 without fixing its position until the end of the same 

 month. WTiUe it was visible with difficulty to the naked 

 eye, in a i6-feet telescope it appeared " en forme d'une 

 petite dtoile n^buleuse avec une chevelure autour d'elle, 

 dont r^tendue paraisait au moins aussi grande que le 

 diametre de Jupiter, vu avec une pareiUe lunette."' This 

 is about the extent of our information relating to the 

 aspect of the comet 



The best orbits are those of Burckhardt (hyperbola and 

 parabola), published in the " Coimaissance des Temps " 

 for 1 82 1. If we employ the parabolic elements we find 

 the following distances of the comet from the sun and 

 earth expressed in units of the earth's mean distance from 

 the sun :— 



