53 



NA TV RE 



Nov. i 8, I SSo 



Nilson and Pettersson have recently repeated their 

 determinition of the specific heat of beryllium, and find 

 these numbers : — 



o°- 50° spec, heat — 0*3973 : o°-ioo° spec, heat = o'\2\(>, 



o°-20o° „ = 0-475 : o°-300° „ = 0-5055. 



If the atomic weight is taken as lyd then the atomic heat 

 for the interval — 



o°-5o° = 5-46 : o°-ioo' = 5-79 : o°-200° = 6-48 : o°-3oo° = 6-9, 

 hence the Swedish chemists conclude that the atomic 

 weight of beryllium is I3'6. 



But ir. the Ust number of the Berichte of the German 

 Chemical Society, Lothar Meyer has calculated, from 

 Nilson and Pettersson's numbers, the true specific heat 

 {i.e. the ratio between the quantity of heat required to 

 raise unit weight of tlie given substance through 1°, start- 

 ing from the given temperature, and the quantity of heat 

 required to raise unit weight of standard substance through 

 1°, also starling from the given temperature) of beryllium 

 for various temperatures : his results are as follows : — 

 (7 = true specific heat at temperature / : A 7 = value of increase 

 of specific heat for 1°). 

 t. t. t. 



+ 23° 7 = 0-3973 ... 73-2 7 = 0-4481 ... 157° 7 = 0-5193 

 A7 = o'ooioi ... A7 = 0000S5 ... A7 = 000063 



256 -S 7 = 0-5319. 

 Hence the atomic heats of beryllium are : — 



<■ Be = 91. Be = 13-65. 



23° 3-62 5-43 



73 4'oS 6-12 



'57" 473 7'io 



257 529 S-94 



The value of Ay decreases as the temperature rises ; in 

 this respect beryllium resembles boron, carbon, and 

 silicon. For other elements whose specific heats increase 

 with increase of temperature the value of Ay also in- 

 creases. Lothar Meyer therefore concludes that beryllium 

 is analogous to boron, carbon, and silicon, in that its 

 specific heat increases as temperature increases, and in 

 that the value of this increase is less for 1° at high than 

 at low temperatures. Hence the atomic weight of beryl- 

 lium is almost certainly 9-1, the o.xide is BeO, and the 

 rnetal finds its place in Mendelejeft"'s system of classifica- 

 tion of the elemsnts according to their atomic weights. 



THE PHOTOPHONE 

 ■[\/rANV readers of Nature will doubtless be glad to 

 -'■•-'- know that Mr. Graham Bell's extraordinary experi- 

 ments may be repeated on a small scale with very simple 

 apparatus, no special appliances being required beyond 

 the mirror transmitter and the selenium receiver, both of 

 which may be easily constructed. I propose to give a 

 short description oi' an arrangement which has m my 

 hands been very successful. 



The mirror is made of the thin mica which is sold by 

 opticians for covering carte tie visile photographs. It is 

 cut by scissors into a circle 2} inches in diameter, and 

 silvered by the process for silvering glass specula. The 

 box m which it is mounted is an ordinarv wood turned 

 box 2i inches in diameter. A circular hole of about 2 

 mches diameter is cut in the lid, behind which the mirror 

 IS laid wuh the reflecting side outwards, a flat ring of vul- 

 canised india-rubber of suitable size and thickness bein^ 

 placed behind the mirror ; when the box is closed tlic 

 ring should hold the mirror firmly in position. If the 

 hd screws on, so much the better. At the bottom of the 

 box IS cut a hole, into which is glued one end of a flexible 

 speaking-tube iS inches long, having at its other end a 

 wooden mouthpiece. It will be found convenient to 

 attach a short wooden arm to the box in a direction per- 

 pendicular to its axis. By means of this arm the trans- 

 mitter may be held in a clamp in any desired position. 



This completes the transmitter as described by Mr. Bell. 

 I have made a small addition which, though not essential, 

 is a decided improvement. At tlie back of the mirror I 

 cemented a disk of calico i inch in diameter, in the centre 

 of which had been previously inserted a loop of silk half 

 an inch long. A hole J inch diameter is bored perpen- 

 dicularly in the side of the box at a point about \ inch 

 from the mirror end of it, and in this hole is inserted a 

 piece of watch-spring ig inch long, with its flat sides 

 parallel to the top and bottom of the box. The spring is 

 fixed into the hole with wooden plugs so that one end is 

 flush with the outer surface of the box ; the other end 

 where it intersects the axis is bent into a shallow hook. 

 Into this hook is slipped the silken loop, and the tension 

 of the spring draws the mirror into a slightly concave 

 form, and seems to make it respond more perfectly to 

 sound vibrations. 



By far the most important part of the whole apparatus 

 is the selenium "cell." After making some dozens of 

 different forms, most of w^hich were more or less sensitive, 

 but none satisfactory, I tried the one now to be described, 

 which turned out very successful. Take a slip of mica 

 2j inches long and ;,' inch broad, and beginning at \ inch 

 from one end, wind round it in the form of a flat screw 

 some No. 40 copper wire. The pitch of the screw is 

 I'j inch, that is, each wire on the two faces of the mica 

 is yV inch from its neighbours. Continue winding up to 

 :r inch from the other extremity; then fix the two ends of 

 the wire by passing them through holes drilled in the 

 mica. Now take a second wire and carefully wind this 

 on beside the other, thus forming a second screw, the 

 threads of which are midway between those of the origi- 

 nal one. Fix this as before. Great care must be taken 

 that the two wires do not touch each otherat any point : it 

 will be well to make sure of this by testing with a galvano- 

 meter before proceeding further. If a lathe is at hand, 

 the tedious operation of winding may be very greatly 

 facilitated. Turn a cylinder of hard wood 4j inches long 

 and I inch in diameter : cut this cylinder longitudinally 

 into two equal parts, and between the two semi-cylinders 

 thus formed place, sandwich-like, a slip of mica of equal 

 breadth. Secure the ends with screws. Smooth down 

 the whole in the lathe, and when the edges of the mica 

 are quite flush with the surface of the wood, cut upon the 

 cylinder a screw- of thirty-two threads to the inch. On 

 removing the mica from the cylinder its two edges will be 

 found to be beautifully and regularly notched. Wind the 

 first wire into alternate notches, and the second into the 

 others. The wire should be annealed to take away its 

 springiness and make it lie flat, and the mica should be 

 stout enough to bear tight winding without buckling. 



For the succeeding operation a retort-stand at le.ist 

 15 inches high is convenient. Fix one ring 15 inches 

 above the foot ; on a lower ring stand a medium-sized 

 Bunsen burner. On the top ring lay a flat sheet of brass 

 ]\ inch thic!:, and on the brass a piece of mica (to save 

 waste selenium). Place the embryo cell on the mica, laying 

 small weights on its two ends to keep it steady and bring 

 it into closer contact. Having brought the Bunsen burner 

 close under the brass, melt a few grains of vitreous sele- 

 nium in a small spoon and let four or five drops fall upon 

 diflerent parts of the cell- Spread the melted selenium 

 evenly over the surface w'ith a slip of mica, pressing it 

 well between the wires. During this process the tem- 

 perature must be carefully regulated by raising or depres- 

 sing the burner. If it is not high enough, the selenium 

 will begin to cr)'stallise ; if too high, the selenium will 

 gather up into drops, being apparently repelled from the 

 surface of the cell. The temperature should in fact be 

 just above the fusing point of crystalline selenium. When 

 a smooth surface is obtained, quickly remove the cell with 

 microscope forceps and let it cool. Its surface will now- 

 be smooth and lustrous. 



The cell must next be annealed. And here my expe- 



