582 



NATURE 



[Oct. 20, 1 88 1 



simple to Stokes and Tliomson in 1852, and to Stewart and 

 Kirchhoff a few years later. 



I wish to consider briefly, what are these new and puzzflng 

 complications of the ^olar problem ; and whether we may not 

 still preserve our belief in the existence of essentially different 

 elementary atoms, which is the basis of the beautiful Vortex 

 Theory. For it seems that to hazard (however naturally) such 

 a step as is involved in assumed dissociation of the (so-called) 

 elements, before we make certain that no less serious hypothesis 

 will account for the observed facts, is contrary to the spirit of 

 Newton's ReguUe Pliilosophandi. 



The most prominent of these complications seem to be — 

 (l). The variations of the relative brightness, width, &c., of 

 the lines in the spectrum of a particular substance, in dependence 

 on the source and circumstances of its incandescence. 



(2). The so-called "long" and "short" lines. (These, as 

 will be seen, are probably a case of (i).) 



(3). The fact that, in the spectra of sun-spots, some lines sup- 

 posed to be due to a particular element indicate rapid motion of 

 the' glowing gas ; while others, supposed due to the same element, 

 give no such indication. 



(4). The (at least apparent) coincidence of lines in the spectra 

 of two or more elementary substances. 

 To these may be added :— 



(5). The remarkable peculiarities of star-spectra j especially 

 the paucity, and the breadth, of the lines in the spectra of white 

 stars. 



As regards (i), let us consider a sounding body with a large 

 number of different modes of vibration, exposed to impacts 

 either periodic or at least with an average period. The reladve 

 intensities of the various notes which it can give will obviously 

 depend upon the period of the impacts. Now this is preci-ely 

 the case of a particle (I use the word to avoid misconception) of 

 a glowing gas. The average number of bio .vs it receives will 

 depend on [a] the number of panicles per cubic inch (and also 

 upon whether there be another gas present or no, a point of very 

 great importance) and (i) the temperature, which is directly 

 connected with the velocity of the particles. 



Change the density, the temperature, the admixture with 

 foreign substances, or any two, or all, of these ; and the average 

 period of the haltering to which a particle is subjected may be so 

 altered as to elicit from it in any required ratios of relative 

 intensity the various simple rays it can give out. 



It will readily be seen that this may account for all of the 

 phenomena of classes (i) and (2) above. 



(3) may be accounted for in many ways. I mention only one, 

 as my oliject is merely to .show that we are not yet compelled to 

 accept dissociation of so-called elements even in its mildest 

 form. Other modes of escape, though not quite so simple, 

 present themselves. 



What is seen in a sun-spot is the integral, as it were, of all 

 that is taking place (as regards both radiation and absorption) in 

 many thousand miles of solar atmosphere, containing the same 

 suSstance under the most varied conditions. That portions in 

 which certain lines of that substance are prominent over others 

 may be at rest relatively to the observer along the line of sight ; 

 while others, in which (from different density, temperature, or 

 admixture, as above explained) other lines are specially prominent, 

 may have large relative velocities, is certain. This would at 

 once account for these singular observations. 



As to (5) we must remember that in a star spectrum we have, 

 as it were, a triple integral. For we not only integrate through 

 the depth of the atmosphere, lout also over the whole surface of 

 the star ; spots, hurricanes, and rotation of the whole, included. 

 This is equivalent to the superposition of innumerable separate 

 spectra, no two of which may have any one individual line in 

 the same place or of the same breadth, &c. Feeble lines may, 

 in fact, entirely disappear under such treatment. 



(4) If not due to want of dispersive power in the apparatus, 

 this may be legitimately attributed to inevitable impurities. It 

 is only in "tall talk" (or in advertisements) that any human 

 preparation, elementary or not, can be spoken of as " chemisch 

 rein." And we all know how faint a trace of impurity can be 

 detected by the help of the spectroscope. 



Even in the last resort, I see nothing to hinder the existence 

 of exactly equal vibration-periods in two perfectly distinct vortex- 

 atoms :— though their occurrence is extremely improbable. 



If we could get an absolutely transparent gas ; one, therefore, 

 which could give no radiation under any circumstances ; the study 

 of the behaviour of a given qu.mtity of hydrogen mixed with dif- 



ferent proportions of it in a vessel of given size, and subjected 

 always to the same conditions of incandescence, would give us 

 invaluable information. G. H. 



Replacing Flakes on Palseolithic Implements 



This wonderful feat was first performed by my friend Mr. 

 F. C. J. Spurrell of Dartford. On first thoughts the thing 

 seems utterly impossible, and it is obvious that no flake can 

 possibly be replaced upon an implement unless one lights on the 

 exact spot where the instrument was made, and finds both imple- 

 ment and flakes in position Mr. Spurrell so found his m.iterial. 

 During the pre ent summer I have discovered another and similar 

 Puloeolithic floor, far removed from Mr. Spurrell's, and where 

 implements and flakes are exposed in a stratum perfectly undis- 

 turbed since they were gently covered up in Palaeolithic times 

 with fine sand containing the shells of such freshwater molluscs 

 as Unio, Cyrenia, and Bythinia. For obvious reasons — the 

 chief one of which is that my work would be totally stopped if 

 I mentioned the locality — I will content myself with stating that 

 the position is nearly a mile from any river, and \\\i floor is 41 

 feet above thi level of the nearest stream ; above the floor is a 

 thick deposit of fine stratified sand, and above that loam. On 

 this Palisolithic floor I have found several implements and a 

 large number of flakes, and on one of the finest implements, 

 an example 5 inches long, 3I inches wide, and weighing \\ lb. 

 I have been able to replace two flakes, one 2\ inches long, the 

 other 2 inches in exact position ; the flakes slightly overlap 

 each other on the implement, and both have been st.uck from 

 the edge of the implement at right angles across its face. The 

 iin;.)lement and flakes were clo^e together, and with them I 

 found a hammer-,-tone of flint with a distinctly battered and 

 abraded edge. Mr. Spurrell replaced many flakes round his 

 implement, but the implement itself was a spoilt and poor 

 e.xample. My implement, on the contrary, is an unusually fine 

 one, large, heavy, and perfect. Both the implement and flakes 

 show a little of the original grey crust of the flint from which 

 the instrument was made, and this peculiar grey colour led me 

 to attempt the replacement of the flakes with the above-men- 

 tioned successful result. One flake has a slightly uneven edge- 

 in some instances considered a proof of u e — the sec nid flake is 

 quite sharp. I shall exhibit this implement, with other imple- 

 ments, flake;, &c., from the same place, at an early meeting of 

 the Anthropological Institute. Worthington G. Smith 

 125, Grosveno;- Roal, Highbury, N. 



Integrating Anemometer 

 Perh.\ps the foUoaing brief description of the integrator 

 devised by me will suffice to e-tablish its near kinship with Mr. 

 Wilson's (Nature, voL xxiv. pp. 467 and 557) : — A roller with 

 a spherical edge is made to revolve with a velocity proportional 

 to that of the wind as recorded on an anemogram. This roller 

 presses on a plane table carried by two mutually perpendicular 

 pairs of rails in planes parallel to that of the table. The lowest 

 of the pairs of rails is supported by a frame carried on the 

 extremity of a vertical shaft. The point of contact of the roller 

 with the table lies in the prolongation of the axis of the shaft. 

 The table can rotate with the shaft, but not independently. By 

 a simple arrangement the shaft, and consequently the table, are 

 caused to take up positions corresponding from moment to 

 moment with the direction of the wind record on the anemo- 

 gram. A style concentric with the shaft presses lightly against 

 a compound sheet of tracing and carbonised paper attached to 

 the under side of the table. Arrangements are also made 

 for obtaining the sum of the movements of the table toward each 

 of the four cardinal points. If the roller be moved with a 

 velocity proportional to that of the wind, whether directly by a 

 cup-anemometer or by a mechanical translation of the trace as 

 given by such an instrument, while the table simultaneously 

 assumes orientations corresponding to the direction of movement 

 of the air, the line drawn by the style will be a miniature copy 

 of the path of an imaginary particle animated by the movements 

 actually belonging to the masses of air which successively affect 

 the anemometer at the given station during the selected period, 

 rigorously in accordance with the principle knosvn as Lambert's. 

 But in order that the trace drawn as described should correctly 

 represent the actual movements of the air, it is evident that the 

 whole mass of the atmosphere must be supposed to move 

 " parallel to itself," i.e. in such a manner that the straight 



