428 



NA TURE 



[February io, 1910 



errors of the elements are of the same order as the per- 

 turbations by Jupiter. 



These perturbations by Jupiter depend upon the exact 

 circumstances of the approach to Jupiter, and these circum- 

 stances in their turn may be largely modified by changes 

 in the elements quite small enough to be consistent with 

 the three months' observations in 1S89-90. There is, 

 therefore, great uncertainty as to where comet Spitaler 

 should now be, and also some uncertainty as to what are 

 the true elements of comet igoge. 



Identity is therefore very far from certain. The excuse 

 for putting forward the present_ conjecture is the interest 

 that naturally turns upon the question of what becomes of 

 the short-period comets that are only once seen. 



P. H. Co WELL. , 



Pleochroic Hales. 



In a recent reference to the subject of pleochroic halos 

 {Phil. Mag., February, igio) I stated that the outer edge 

 of a corona might present an appearance suggesting an 

 actual accentuation or deepening of the coloration, in 

 accordance with the fact observed by Bragg and others that 

 the ionisation of the o ray increased just before the limits 

 of its ionising range was attained. For certain stages of 

 development of the halo, this observation I have more 

 recently confirmed beyond doubt. Like other structural 

 particulars referred to in my paper, this too becomes 

 obliterated by over-exposure. In one case (in the lithia 

 mica of Zinnwald) a stage of development has been found 

 in which the extreme outer border of the corona is the 

 sole visible part of that structure, the appearance presented 

 being that of a detached, very delicate, ring of perfect 

 regularity surrounding the central halo, a space showing 

 no definitely visible ionisation intervening. The outer ring 

 has a radius of 00344 mm., and the inner halo a radius 

 of o-oigi mm. The outer ring is of about normal radial 

 dimensions ; the inner radius is that corresponding to 

 under-exposure to the slower moving o particles. Refer- 

 ence to Bragg's curves (Phil. Mag., September, 1905) will 

 more fully explain. 



In the granite of Ochsenkopf, Fichtelgebirge, complex 

 halos will be found very beautifully developed. Some of 

 these conform to dimensions such as might be referred 

 to the a radiations of thorium and its derivatives, others to 

 those of the radium family. 



The halos described in my paper, referred to above, are 

 for the most part in a lithia-bearlng mica, of a kind which 

 is not correctly included among the Muscovites. The 

 emendation does not, however, notably affect the calcula- 

 tions given. The careful observation of the dimensions 

 of pleochroic halos will, I think, be found of service in 

 distinguishing certain micas — the Biotites from the Musco- 

 vites, for instance. J. Joly. 



Trinity College, Dublin, January 31. 



Dangerous Lecture Experiments. 



The explosion referred to by Mr. Power in N'.vtl're of 

 February 3 (p. 3qg) was probably due to the presence of 

 a trace of moisture in the reacting substances. I had a 

 similar alarming experience some years 'ago, using pre- 

 cipitated silica without specially drying it. 



Moissan (" Traits de Chimie MinJrale," ii., 389), re- 

 ferring to a paper by Ludwig Gattermann (Ber., xxii., 186, 

 i88g), states: — "La r(^action (SiO, + 2Mg = Si + 2MgO) est 

 si violente, d'apr^s cet auteur, que si Ton emploie la 

 silice pr^cipit^e, le tube de verre est enti^rement d^form^, 

 et une partie le la mati^re est projet^e sous forme d'une 

 gerbe de feu." Winkler (Ber., xxiii., 2652, iSgol found that 

 0-2 gram of a mixture of magnesium and silica in the 

 above proportions heated in a tube closed at one end caused 

 explosion and shattering of the tube. Vigorou.x (Annales 

 de Chimie ct de Physique, xii., 153, 1897) recognised that 

 the explosion is due to incomplete desiccation of the react- 

 ing materials. 



Few text-books point out the necessity for ensuring 

 the absence of moisture, although most of them point out 

 that the reaction is very rapid and violent. The only book 

 besides Moissan's " Traits " that I have found to give 

 the warning is by an .American, Dr. Benedict (" Chemical 

 Lecture Experiments," The Macmillan Company, New 

 NO. 2102, VOL. 82] 



York, 1901). Dr. Benedict insists upon the absence of 

 moisture, but does not mention any reason. 



This is not the only case in which dangerous experiments 

 have been described and copied from one text-book to 

 another. The collection of hydrogen from the action of 

 sodium upon water is a case in point. Many books 

 describe, with a diagram, the " drowning " of a piece of 

 sodium by means of a special instrument of wire gauze. 

 This 7nay be carried out safely in some instances, but, as 

 \ewth points out, it is liable to be dangerous. Lfpon one 

 occasion in my own experience, using a small piece of 

 sodium, an earthenware pneumatic trough was shattered 

 as well as the gas jar used to collect the hydrogen. 



It would seem desirable that writers of te.xt-books should 

 obtain some personal knowledge of the experiments they 

 recommend, as young teachers, relying upon the instruc- 

 tions given, might easily cause very dangerous accidents. 

 In neither of the cases cited above would a chemist, unless 

 fairly experienced, be likely to apprehend any difficulty or 

 danger. E. R. Marle. 



Hartley University College, Southampton. 



The letter of Mr. Power in Nature of February ' 3 

 directs attention to a danger common to the use of any 

 of the metallic reducing agents, and, although well known 

 to a few for many years, it is not at all generally recog- 

 nised. Even the well-known reaction 



Fe.O, -I- 2.^1 = .'VUOj -I- 2 Fe, 

 if performed in the way described, is positively dangerous 

 unless all the precaution necessary for a violent explosion 

 be taken. 



A very striking lecture experiment is to charge a tiny 

 steel crucible that will contain about one-half up to one 

 cubic centimetre of a mixture of ferric oxide and finely 

 powdered aluminium, and to cover it loosely with a thin 

 sheet-iron cover so as to preserve the contents from water 

 vapour, and then to heat this up in a little furnace made 

 of strong iron gauze covered with asbestos and held rigidly 

 in a retort stand. .\n ordinary blow-pipe with a foot 

 blower will be sufficient, and the reaction is so violent, as 

 soon as the necessary temperature is attained, that in 

 nearly every case the steel crucible will be shattered into 

 pieces, notwithstanding the lightness of the cover. 



The possibly dangerous character of the reduction was 

 noticed by me certainly so long ago as 1S96 whilst pre- 

 paring special qualities of iron experimentally in the South 

 Kensington laboratories, and shown to many persons. 

 Even then it may have been not unknown to other workers 

 with metals, although new to us ; and although the mix- 

 ture was at once respected, and absorbents of heat were 

 used in the charge to moderate the action, I am aware of 

 at least one narrow escape by an operator who wished to 

 verify the observation and used a quarter of a kilogram of 

 the mixture in a crucible heated by a Fletcher oxygen 

 injector furnace, the pieces of which were thrown all over 

 a large room, fortunately without striking any person. 



The explanation is simply that the preliminary heating 

 to the temperature of reaction is sufficient to en.ible the 

 mixture to reach the volatilisation point of the iron by 

 the heat suddenly evolved throughout the mass, and thus 

 there is practically detonation ; but it also suggests that 

 some danger of explosion exists should a store of the 

 mixture be involved in a fire, and these mixtures are now 

 in fairly common use industrially. 



Henry C. Jexktxs. 



The School of Metalliferous Mining (Cornwall^, 

 Camborne, February 5. 



The Maintenance of Forced Oscillations, 



Please permit me to add a few words to my note on 

 " The Maintenance of Forced Oscillations of a New Type," 

 which appeared in Nature of December 9, 1909. I stated 

 that when a vibrating fork maintains the vibration of a 

 string by periodically varying its tension, the stationary 

 oscillation maintained may have a frequency of half of, 

 equal to, 3/2 times, twice, &c., of that of the fork, each 

 term in the harmonic series appearing separately by itself, 

 or with one or more of the others conjointly, according to 

 circumstances. 



When two or more of the harmonics thus appear cQn- 



