558 



NA TURE 



[April 1 1 , I 



ably on account of absorption in the cloud regions of our atmo- 

 sphere, which, as Langley has shown, takes up with great avidity 

 ihe violet and ultra-violet rays. 



May it not be that in clouds we have conditions especially 

 favourable to the production of the Hertz effect ? If so, the 

 discharge from one cloud to another would be accompanied by 

 rn earth current in the opposite direction, as in the theory pro- 

 loosed by Pr..f. Stokes, in which a decrease of resistance is 

 ]iroduced by an increase of heat from the sun. 



Hertz found his effect [IVicd. Ann., xxxi. p. 993) much more 

 marked in a medium under diminished pressure. 



Under 300 millimetres of mercury, he finds that the ultra- 

 violet radiation will nearly quadruple the length of spark ob- 

 lained without it, while under ordmary atmospheric pressure it 

 would scarcely dottble it. But this is the very circumstance which 

 is realized in the ca«e of cl uds. 



There is also reason to think that solar outbursts are especially 

 rich in these rays of short wave-lengih which are required to 

 explain the phenomena. Henry Crew. 



Haverford College, U.S.A., March 22. 



Hertz's Equations in the Field of a Rectilinear Vibrator. 



Recurring to Hertz's equations for the field of the recti- 

 linear vibrator, it appears to me that, while his conclusions are 

 sound as regards the forces at points very distant from the 

 vibrator, they require modification for the rest of the field. In fact, 

 the principles upon which the question is investigated require 

 that the electromotive force in the direction of z should become 

 evanescent close to the vibrator ^the axis of z). 



The general form of n is either — 



M sin ^ 

 A. 



M cos ^ 



sin nt, or 



.sin tit, 



where A large, and Xn = - , or, of course, the sum of the two 



A 

 forms. 



In assuming for points near the origin (say the middle point 

 of the vibrator) the approximate expression — 



M . , 



Hertz, in point of fact, takes the second of the above forms for 



n, for this reduces to — . sin nt when - is very small. 

 A. A -^ 



But this assumption makes both 11 and Z infinitely great close 

 to the vibrator. Whereas, by assuming the former of the two 

 forms, or — 



M sin ?- 



i.e. near the origin n == -- sin nt, we get, as a general 



ejsin, 



expression for Z — 



■ -^ ) sin ^- - 1 — - ?" - I cos 

 t VP" p" A'p- K-pj \ \\p' \p*J 



and, as p is indefinitely diminished, this reduces to — 



2 M . , 

 sin 7it 



3 A' 

 as a limiting value. 



For distant portions of the field. Hertz's results as to the 

 laws and amplitudes of the forces electric and magnetic remain 

 unaltered. 



Of course, the whole investigation, with such a simple 

 assumption as to the nature of the field, must be regarded as 

 only approximate. For any given form of vibrator — as, for 

 example, a straight wire connecting two spheres — the exact 

 treatment will be very difficult. In the simplest conceivable 

 case of a spherical metal sheet with an induced Q„ distribution 

 f ft to itself, the analysis is intricate (see a paper by Prof. J. J. 

 Thom.on to the Mathematical Society of London, January 

 1884). II. w. Wj^'so^. 



Berkswell, March 29. 



Early History of Lightning-Conductors. 



Can any of your readers refer me to the sources of some 

 of the late Mr. Richard Anderson's information with regard 

 to the early history of the lightning-conductor? (i) On p. 



27 of the third edition (1885) of his book, " Lightning-Con- 

 ductors," he states that Franklin, in the 1758 issue of "Poor 

 Richard," gave directions for the erec'.ion of lightning-conductors. 

 (2) On p. 25 he refers to Prof Winthorp, of Boston, having, in 

 1755, defended the lightning-conductor against a parson who 

 had attributed a Massachusetts earthquake to the innovation. 

 I should be much obliged for any reference to a library where 

 a copy of " Poor Richard " for 1758 could be found ; or, again, 

 for any information with regard to Winthorp's defence of the 

 lightning-conductor. 



Prof Meidinger, of Karlsruhe, who is preparing a second 

 edition of his " History of Lightning-Conductors," is extremely 

 desirous of verifying these details of their early history, and I 

 should be glad if any of your readers could supply me with 

 information for him on these points. Karl Pearson. 



University College, April 9. 



The Satellite of Procyon. 



Mr. J. M. Barr's suggestion (Nature, March 28, p. 510), as 

 to the use of photography to ascertain whether there is any close 

 companion or satellite to Procyon, would be considered a very 

 desirable one by astronomers, in order to set at rest the question 

 whether a companion can actually be discovered near the 

 assumed place of the hypothetical one, of which the elements 

 were given by Ur. Auwers in 1861, from investigations of the 

 irregularity in the proper motion of Procyon observed by Bessel 

 in 1844, and by Madler in 1851. The orbit was computed on the 

 assumption of a circular motion in a plane perpendicular to the 

 line of sight round a point about i"'2 distant, having a period of 

 about 40 years, the position angle for 1873 being about 90°, so 

 that the present angle would be about 234", or about 9° per 

 annum. 



I fear, however, serious instrumental difficulties would arise 

 in obscuring such a brilliant object as Procyon in a large tele- 

 scope by a screen, so as to get the impress on a plate of a 

 probably faint companion at the extremely close distance of two 

 to three seconds of arc. 



This difficulty, no doubt, must have presented itself to the 

 minds of the astronomers at the Lick Observatory, California, 

 or they would have tried the sensitive plate with the 34-inch photo 

 lens of the great refractor, instead of examining Procyon visually 

 with the 36-incb glass, as was done by Mr. S. W. Burnham on 

 the early morning of November 18 last, with the following 

 record : — " Frocf on.— CareMly examined with all powers up to 

 3300 on the 36-inch under favourable conditions. Large star 

 single, and no near companion." 



If this means [that no companion was seen within 10" or 12" 

 radius, it makes the matter very perplexing, as Otto Struve 

 measured a supposed new companion in 1873 with the 15-inch 

 refractor at Pulkowa— the mean of several measures for March 



28 being P. angle 90°'24, and distance I2"'49, and for 1874 

 (April 10) P. angle 99° -6, and distance ii"-6-/. This companion 

 was looked for at Washington with the 26-inch refractor on 

 several occasions from November 1873 till January 1876, and by 

 the three Clarks (father and two sons) with the McCormick 

 26i-inch refractor, then completed at Cambridgeport, Massa- 

 chusetts, but Struve's companion could not be seen with either in- 

 strument, and I am not aware that it has since been seen by 

 Struve himself with the new Russian 30-inch refractor. The 

 Washington observers at that time, however, gave estimated 

 places for three new companions, supposed to be seen by them 

 as follows : — 



No. I. Position angle, about 10°, and distance about 6". 



.. 2. „ „ „ 36° „ „ „ 8"-8 



M 3- ,. -. ,, 50° M „ ., 10" 



These appear (if they have an existence at all) to have been 



missed with the 36-inch glass at the Lick Observatory, as above 



referred to. 



It is a singular coincidence that the position- angles of the 

 companion supposed to have been seen by Otto Struve in 1873 

 and 1874 agreed with the orbital places computed by Dr. 

 Auwers, but its distance involved the assumption of an enor- 

 mous mass to Procyon for the parallax assigned to the principal 

 star. Isaac W. Ward. 



Belfast, April i. 



