Nov. 1 8, 1 8 So] 



NA TURE 



59 



lience differs in a remarkable manner from ihat of Mr- 

 Bell, as stated in his celebrated lecture. It is true that 

 selenium may be rendered crystalline in " a few minutes," 

 but in this condition I find it far less sensitive to light 

 than after it has undergone a process of long heating and 

 slow cooling. My method is as follows: — The brass 

 plate being cool, lay the cell upon it again, and place the 

 burner at its lowest possible point. The selenium will 

 soon begin to crystallise, as evidenced by its surface 

 assuming a dull leaden appearance. (If the crystallisa- 

 tion has not begun in five minutes, raise the burner an 

 inch or two.) In from five to ten minutes the whole 

 of the selenium should be crystallised. Then very gra- 

 dually raise the burner until signs of fusion just begin to 

 appear. This will probably take place when the flame is 

 within 3 inches of the brass. Instantly remove the 

 burner, and in about ten seconds re-crystallisation will 

 occur. Now fix the burner \ inch below the point at 

 which it was when fusion commenced, and let it remain 

 for four hours, merely looking at it from time to time to 

 ascertain that, owing to increase of gas pressure or other 

 causes, the heat has not become too great. After four 

 hours begin the cooling by lowering the burner an inch 

 or two, and repeat this operation every ten or fifteen 

 minutes, until the burner is at its lowest point. Then 

 slightly lower the gas-flame at short intervals, until it is 

 finally extinguished. When the brass plate is quite cool 

 the cell may be removed. 



I may mention that I first made a cell of this form, 

 which I believe to be original, on October 28.^ If the 

 two wires were wound on a cylinder made of some 

 suitable non-conductor {c.f;. slate) with a double screw 

 cut upon its surface, a cell might be formed which, it 

 appears to me, would unite all the advantages of Mr. 

 Bell's with far greater simplicity. 



My experiments were made with the transmitter and 

 selenium cell above described, a magic-lantern with a 

 4-inch condenser, the focussing lenses being removed, 

 two plano-convex lenses obtained by separating a 3i-inch 

 condenser, a "blow-through" lime-light, a battery of 

 eleven cells (small Leclanchc's answer well), and a pair 

 of Bell telephones. It is essential that the bobbins of the 

 latter be wound with finer wire than that generally 

 used. Mine contain No. 40 (instead of 35 or 36), and I 

 intend to try 43. Their diameter is also larger than 

 usual — 1| inch. 



The transmitter is clamped so that its axis is inclined 

 at an angle of about 30° to that of the lantern condenser, 

 the centre of the mirror being 7 or 8 inches from the 

 centre of the condenser ; and the position of the lime- 

 light is so adjusted that the condensed rays may just 

 cover the whole surface of the mirror. 



The reflected beam is rendered as nearly parallel as 

 possible by one of the plano-convex lenses (this can only 

 be done approximately), while the other, placed a foot or 

 two away, concentrates the light upon the selenium cell, 

 forming an elliptical image of the mirror. The mjjor 

 axis of the ellipse should be parallel to the length of the 

 cell, and the minor axis slightly longer than its width. A 

 great deal depends upon the focussing, and the best 

 results have been obtained when the image of the mirror 

 was not quite sharp. The selenium cell is joined in circuit 

 with the battery and the pair of telephones, the latter 

 being for obvious reasons placed in a distant room. The 

 arrangements are now complete, and a person listening 

 with a telephone applied to each ear w-ill, if everything is 

 right, plainly hear words which are spoken into the trans- 

 mitter. When I first made the experiment I was so 

 much astonished at the distinctness of the reproduction 

 that I believed that one of the battery connections must be 



' If a larger surface is desired, two or more cf these cells may be placed 

 t5<:cther side by s'.de, the ends of the wires being properly connected. The 

 width of \ inch for a single cell cannot be much exceeded, because the 

 expansion pri^duced by the heat necessary for melting the selenium would 

 in.ike the wires on a wider surface so loose as to touch each other. 



defective, thus acting like a microphone. This was dis- 

 proved by screening the mirror, when all sound instantly 

 ceased. 



Though the articulation is not perfect, it is far better 

 than I had expected, judging from the accounts of the 

 performances of the photophone in Paris. A leading 

 article might not be altogether intelligible, but ordinary 

 colloquial phrases are readily [understood. The loud- 

 ness of the reproduced speech varies in an unaccountable 

 manner. Sometimes the voice is rendered almost as loudly 

 as in an ordinary telephone ; at other times, under ap- 

 parently the same conditions, it is scarcely audible. Alter- 

 nations from loudness tofaintness,and vice vcrsA, frequently 

 occur in a single sentence. 



The distances across which the beam is carried have 

 varied in my experiments from i foot (when the two 

 plano-convex lenses were in actual contact) to rather 

 more than 4 fcet.i With a larger receiving lens this 

 distance could be greatly extended, especially if the 

 electric light were used. 



For the " musical" effects produced by an interrupted 

 beam I use a disk of zinc i foot in diameter, having eight 

 radial slits cut in it, and mounted upon a vacuum tube 

 rotator. The cell is placed 6 inches from the lantern 

 condenser, and the disk made to rotate close before it. 

 The sound produced is very loud, and can be heard when 

 the telephones are at a distance of a foot or more from 

 the ears. 



It is very singular, that whereas I have been so suc- 

 cessful in repeating Mr. Bell's more complex experiments, 

 I have utterly failed in all attempts to produce sound by 

 the simple incidence of an interrupted beam upon a thin 

 diaphragm. I have experimented with disks of ebonite 

 varying from ,}^ to \ inch in thickness, and with several 

 metals, and can only suppose that my source of light is 

 not sufficiently powerful. Shelford Bidwell 



THE CHRONOGRAPH 



MESSRS. E. DENT AND CO., of the Strand and 

 Royal Exchange, London, have been for some time 

 past at work upon three galvanic chronographs of unusual 

 accuracy and po'.ver. They surpass in both respects, so 

 far as we know, any similar instruments yet constructed ; 

 and we believe, therefore, some account of them will be 

 interesting to our readers. They are destined respectively 

 for the Roval Observatory of Brussels, for the Japanese 

 Government, and for the Egyptian Government. 



The advantages of the " chronographic" registration 

 of the times of observations in observatories are not to be 

 o-ainsaid. In the absence of any such arrangement an 

 astronomer, whilst watching through his telescope, has to 

 compute the time by counting up the clock-beats. More 

 often than not he will find that no clock-beat exactly 

 coincides with the instant of his observation. He must 

 then reckon the difference— the fraction of the second 

 elapsed— by judgment as best he is able. Skilled 

 observers can reckon to tenths of seconds, but these are 

 lar'^e and coarse amounts compared with what may be 

 notlid upon such chronographs as those we are referring 

 to In any case the astronomer must make a hurried 

 memorandum of his results ; otherwise he is liable to 



forget them. , ^ . , • 



The Astronomer-Royal was, we believe, the first to in- 

 troduce a system of astronomical chronographic measure- 

 ment into England ; and he designed and had constructed 

 at the Royal Observatory a large apparatus for the pur- 

 pose The reader must bear in mind that though differ- 

 in<^ in some respects both in their mechanism and the 

 means employed, the chronographs we are going to 

 describe are fundamentally the same as the Astronomer- 

 Royal's. 



F > With a sl-inch receiving lens the distance has been increased to upwards 

 of 10 feet. 



