Oct. 



1884] 



NA TURE 



635 



been staying for ten days in London, and two long white 

 paraffin candles have been standing on the drawing-room chim- 

 ney-piece all the time. We have not been using the candles ; 

 but the wicks were ignited before we came to the rooms, as it is 

 very common to do with new candles. I noticed two days ago 

 that the wicks were all covered over with what I at first took to 

 be some kind of mouldy growth, but what I now find is dust 

 which has attached itself in fine hair-like tufts to the wicks of 

 the candles, sticking out in all directions exactly like the tufts of 

 iron filings on a powerful magnet. 



I am well acquainted with this phenomenon in the vicinity of 

 an electric machine which is kept working continuously. For 

 example, we find it constantly in the neighbourhood of the elec- 

 trifying machine of Sir W. Thomson's siphon-recorder, where 

 the insulating supports collect great quantities of dust, and 

 generally in these curious forms. From the appearance of the 

 deposits on the wicks of the candles I have very little doubt that 

 somehow or other electric attraction has played an important 

 part in the collecting of the dust and the formation of the fila- 

 ments. But whence the electrification has come I am unable 

 to say, unless it be that a warm current of air, which I find is 

 always passing upwards past the candles from a wooden mantel- 

 piece warmed by the fire, electrifies the paraffin candles and 

 causes the phenomenon I have described. 



J. T. Bottom ley 



39, Eastbourne Terrace, London, \V. , 

 October 25 



Simson's Line 



Mr. J. S. Mackay of the Edinburgh Academy, though not 

 able to trace " Simson's line " to Simson's works (see my notice 

 of Dr. Casey's "Euclid," Nature, October 23, p. 607), has fur- 

 nished me with the following account, which may be of general 

 interest: — " The theorem that the orthogonal projections of a 

 point on the circumference of a circle upon the sides of an in- 

 scribed circle are collinear is ascribed to Robert Simson by 

 Catalan in his 'Theoremes et Problemes de Geometrie Elemen- 

 taire,' and he speaks several times of 'la droite de Simson.' 

 This book of Catalan's is, I fancy, better known in the United 

 Kingdom than many other Continental works where the same 

 statement is made ; and I conjecture that we have adopted the 

 name from Catalan. It may, however, be the case that we have 

 taken the information from Poncelet's ' Propriete's Projectives,' 

 § 468, where it is said that Servois attributes the theorem to R. 

 Simson. The passage where Servois makes this ascription 

 occurs in Gergonne's 'Annales de Mathematiques,' vol. iv. 

 p. 250, and it is unsatisfactory enough : ' La methode qui vient 

 d'etre indique plus haut pour determiner le point C repose sur 

 le theoreme suivant, qui est, je crois, de Simson.' I cannot 

 ■carry the ascription of the theorem to Simson farther back than 

 to Servois, and though I am not positive that Servois has made 

 a mistake, yet I think it highly probable. The extension of the 

 theorem to the oblique projections is attributed by Catalan to 

 Chasles. It is due to Poncelet, and is given in the section 

 quiited above." The Writer of the Notice 



October 27 



A Rainbow after Sunset 



On the evening of August 29 the almanac sunset for Ireland 

 south is placed at 6h. 51m. Happening to look out to south- 

 south-east I saw a well-marked, though not very brilliant, portion 

 of a rainbow in a shower cloud just above the horizon. It was 

 not a perfect bow, but what sailors call a dog. I looked at the 

 clock and saw that it was 7.15 p.m. Knowing that the sun had 

 set, and being curious to see what could have produced the bow, 

 I immediately went out and examined the western sky. The 

 sun had indeed set, but there was a bright red glow and some 

 ilocculent clouds were tinged strongly with a brilliant rosy red. 

 It was plain that the rainbow was caused by reflected light. 



Stonyford, Co. Kilkenny James Graves 



TROPICAL AFRICAN MOUNTAIN FLORA 



\ VERY interesting collection of plants has been 

 **■ brought to Kew by that intrepid African explorer 

 Mr. Joseph Thomson, made during his late journey into 

 the Masai country. They have been examined by Prof. 

 Oliver, and consist of about thirty-five species from Kili- 



manjaro at 9000 to 10,000 feet of elevation ; a few from a 

 crater near Lake Nairasha at 7000 to Sooo feet eleva- 

 tion ; thirty-four from the Kaptd plateau at 5000 to 6000 

 feet ; and fifty-eight from Lykipia at 6000 to Sooo feet. 



These collections exhibit the mingling of North Tem- 

 perate types with others characteristic of Southern Africa, 

 for which previous discoveries had prepared us. Of these 

 the most interesting are, as new to Tropical Africa, an 

 Anemone,a Delphinium (very different from the Abyssinian 

 D. dasycaulon), and a Cerastium of remarkable habit. 

 Of South African forms the most striking is the handsome 

 arborescent Rutaceous plant, Calodendron capense, the 

 " wild chestnut" of Natal, to the north of which it had 

 not previously been found. Of northern forms is a 

 Juniper, another genus unknown to Tropical Africa, and 

 which was found forming groves at an elevation of 6000 

 to 8000 feet, and itself attaining a height of 100 feet ! it is 

 the J. procera of Abyssinia. A Podocarpus gathered 

 along with the Juniper, and also attaining 100 feet in 

 height, is probably the P. elongata of Abyssinia, which, or 

 a near ally, also occurs in South Africa. The only other 

 Conifer previously found in the equatorial regions of 

 Africa is the Podocarpus Mannii from the peak of St. 

 Thomas in the Gulf of Guinea. J. D. Hooker 



AN ELECTRO-DYNAMOMETER WITH 

 EXTREMELY LIGHT SUSPENDED COIL 



IN my former communications to NATURE it has, I 

 believe, appeared (1) that the induction currents used 

 by Du Bois-Reymond, Duchenne, and other observers 

 for physiological and therapeutical purposes were only 

 arbitrarily and very insufficiently measured ; (2) that the 

 simplest and most practical instrument for their measure- 

 ment is a delicate electro-dynamometer ; (3) that in con- 

 sequence of their extreme smallness, every available 

 method must be employed to reduce the sluggishness of 

 such an instrument without impairing its accuracy ; (4) 

 that an instrument of this character, shown by me before 

 the Physical Society at Oxford in June 18S2, had 

 answered very well, indeed better than a more expen- 

 sive apparatus designed by Prof. Kohlrausch for larger 

 currents. 



It was, however, objected that there is an insurmount- 

 able difficulty in keeping a good contact between the 

 aluminium and silver-gilt wires used in it for suspended 

 coil and suspending wire respectively. 



At the British Association meeting in Montreal I was 

 able to show an improved form of the contrivance, in 

 which this difficulty was surmounted ; and, in addition, a 

 method of damping the oscillations, which, while im- 

 proving the insulation, enabled the weight of the suspended 

 coil, on which the force of the torsion couple depends, to 

 be varied between limits practically infinite. 



The contact difficulty is met by taking a small plate of 

 ebonite 3 mm. by 5 mm. in size, and tapping into it two 

 small gold screws, long enough to project through, and 

 carry two little nuts on the opposite sides. To the two 

 screw heads the ends of the aluminium coil, bent into 

 rings and filed flat, are firmly screwed ; under the two 

 nuts are twisted the ends of the gilt-silver suspension 

 wires ; the nuts are then similarly screwed home. Ebonite 

 is elastic enough to render the junction air- and fluid- 

 proof. 



The second requirement was attained by coiling the 

 aluminium wire on a thin tube of cork, and immersing it 

 in a vessel filled with petroleum oil. Aluminium is about 

 two and a half times heavier than water, nearly three 

 times the specific gravity of this oil ; whereas cork floats 

 on it. Consequently, by properly proportioning the 

 amount of cork relatively to the wire coiled on it, any 

 desired specific gravity from absolute flotation to that of 

 aluminium itself can be obtained. It is even practicable 

 to load the coil, like a Sykes's hydrometer, by dropping 



