4° 



NATURE 



[November 13, 1902 



Three Stars with Large Proper Motions.— M. A. 

 Verschaffel communicates to No. 3824 of the Astronomischt 

 Nachrickten the positions of the stars B.D. + 24°-2439, 242733 1 

 and 24'2733 2 , as recently observed by him at Abbadia, and 

 compares them with the positions given in the catalogue A. G. 

 Berlin B. and brought to 1900 by the corrections for precession 

 and secular variation given in the catalogue, thereby demon- 

 strating the existence of a large amount of proper motion for 

 each star. 



The Pyramid Spot on Jupiter. — Herr Leo Brenner, in 

 writing to The Observatory (No. 324), explains the great dis- 

 crepancies which have appeared between the positions, and 

 velocity, of the " pyramid " spot as determined by himself and 

 as determined by the English observers Messrs. Denning and 

 Phillips. 



He found that the centre of the formation travelled, during a 

 year, at the mean velocity of o°'5 per day, and then Messrs. Den- 

 ning and Phillips recorded that, according to observations made 

 on June 28, it had moved at a mean velocity of nearly 7 'O per day 

 for a period of nine days. This great change of velocity seemed 

 impossible, but Herr Brenner has found a solution to the diffi- 

 culty in the observed fact that it is not one spot that is being 

 observed, but a series of three or four spots, and of these, some 

 are new formations of which Messrs. Denning and Phillips had 

 measured the position as though they were portions of the 

 original spot, thus obtaining the great differences in position 

 noted above. 



Herr Brenner has arrived at the conclusion that neither the 

 markings seen by him during August and September, nor those 

 seen by the English observers on June 28, can be identical with 

 the " pyramid " spot of last year, and these conclusions are 

 strengthened by the observations of Seiior Comas Sola, which 

 were published lately in the Bulletin de la SocUti Aslronomiqut 

 de France. 



Ephemeris for Comet Tempel :! -S\vikt. — In continuation 

 of the ephemeris given in Astronomischt Nachrieliten, No. 381 1, 

 M. J. Bossert now publishes the following ephemeris for this 

 comet. 



llh. M.T. Paris. 



Decl. log r. log <a. 



- 16 6-2 ... 0-1697 •■• °'l59-6 

 -15 26'6 ... 01591 ... 0-1603 



- 14 42-2 ... 0-1486 ... 0'i6o6 



- 13 52-6 ... 0-1383 ... 0-1603 

 -12 57-5 ... o-i2S2 ... 0-1595 

 -11 56-3 ... 0-1184 ••■ 0-1584 

 -10 49-5 ... o - ioS9 ... 0-1566 



- 9 367 ... 01000 ... 01546 



- 8 17-4 ... 0-0917 ... 0-1525 



- 6 52-0 ... 00S43 ... 0-1502 



- 5 21-0 ... 00776 ... 01479 

 Kiel Circular, No. 53. 



THE A UTOMA TIC TELEPHONE EXCHANGE. 



THE object of the automatic telephone exchange is to dispense 

 -*- with the assistance of a third party in making connection 

 between two subscribers. Those who are at all familiar with 

 the complexity of the connections and of the numerous devices 

 needed in a modern exchange having a large number of sub- 

 scribers will realise that to work out a system in which the tele- 

 phone girl is replaced by an automatic arrangement is a matter 

 requiring no little ingenuity, and will, perhaps, not be surprised 

 that the problem has apparently only been attacked successfully 

 on the other side of the Atlantic. The American technical papers 

 have shown that, during the past few years, the construction of 

 automatic exchanges has received considerable attention and 

 that several different systems have been worked out. Some 

 attempts have been made to introduce these into this country, 

 but not with much success ; in fact, until the last year or so 

 England did not afford a promising field for the introduction of 

 automatic telephones unless for small private exchanges. In 

 America, however, matters are different, and, as we have said, 

 descriptions of two or three different systems in actual or experi- 

 mental use have been published. One of these, recently de- 

 scribed in the Scientific American, is noteworthy for the fact 

 that the automatic apparatus at the exchange is operated 



NO. 1724, VOL. 67] 



mechanically so far as possible, the electrical control being reduced 

 to a minimum. Greater trustworthiness, it is said, is obtained 

 by this means, though we should be inclined to think that the 

 wear and tear would also be greater. We do not know whether 

 this, the Faller, system has had any extensive trial as yet. 

 Another system, which we propose to describe briefly, has been 

 in operation in some parts of America for three or four years, 

 and as it is being installed now in several large American towns, 

 and is also being introduced into Germany and England, we 

 may judge that it has proved both trustworthy and economical. 

 In Chicago, an exchange on this system is being constructed with 

 an ultimate capacity of 100,000 subscribers. 



This system is knovyn as the " Strowger " system. We have 

 had an opportunity of inspecting a small model installation repre- 

 senting part of an exchange suitable for 10,000 subscribers, and 

 were struck by the ease and simplicity of its working and its 

 great convenience from the subscriber's point of view. Of 

 course, working a small portion of an exchange under exhibi- 

 tion conditions is one thing and running the complete system 

 continuously, with all the subscribers connected, is another ; 

 but there was little to lead one to suppose that the working 

 under the more arduous conditions of actual service would be 

 any less satisfactory, and indeed the success which has attended 

 the operation of three or four large exchanges in America is 

 direct evidence to the contrary. One of these, at Fall River, 

 Mass., has been in operation for two years and, with an 

 ultimate capacity of 10,000 subscribers, already has 4000 con- 

 nected. Apart from the clerical staff, only five people are re- 

 quired to look after this exchange, and these are said to spend 

 most of their time connecting up new subscribers ; at night and 

 on Sundays the exchange is left to take care of itself. 



We may first of all consider the subscriber's instrument ; this 

 takes no more room than, and looks very much like, an 

 ordinary wall set. There is, however, no magneto ringer, and 

 on the front of the box is a circular metal disc having ten holes 

 on the right-hand side numbered from o to 9 ; below this is a 

 ringing-up push. Suppose a subscriber wishes to call up No. 

 5683 ; he takes his receiver off the hook in the usual manner 

 and, putting his finger in the hole marked 5, rotates the disc 

 until his finger comes against a stop ; he then allows the disc 

 to return to its normal position and repeats the operation with 

 the holes marked 6, 8 and 3 in succession. He is now con- 

 nected through, and if No. 56S3 is engaged, will hear a buzzing 

 in his receiver ; if not, he has only to press the ringing-up push 

 and wait until his call is answered. When he puts back his 

 receiver on the hook, all the connections are restored to their 

 original condition. The time taken to get connected through 

 —or to find out that the number you require is engaged — is 

 considerably less than with the ordinary system, even when the 

 exchange girl replies to your call and connects you up imme- 

 diately, which, as telephone users know, happens but rarely. 



The apparatus at the exchange consists of a number of 

 automatic switches known as "first" and "second selectors" 

 and "connecting switches." The construction of all these is 

 very similar, but is too complicated to describe in detail ; we 

 can only indicate the principle upon which they work. The 

 switch consists of a semi-cylinder, along the axis of which is the 

 switch-arm. This arm can be raised or lowered in ten steps 

 and also rotated so that its contact can be brought up to any of 

 the contacts on the inside of the semi-cylinder ; these are 

 arranged in ten rows of ten contacts each. We may best under- 

 stand the operation of these switches by following out what 

 happens on ringing up, say, No. 5683. Each subscriber has a 

 "first selector" switch of his own at the exchange, and the 

 first movement of the dial on his instrument operates this 

 switch. As he draws down the hole 5 to the stop, a succession 

 of five current impulses are sent along the line, and these raise 

 the central switch-arm to the fifth row up on his switch. This 

 picks out all the subscribers whose numbers begin with 5000, by 

 connecting the caller to the group of " second selectors " 

 corresponding thereto; there are ten connecting or "trunk" 

 lines leading from the first selectors to the second, and the 

 switch-arm, when it has risen to the fifth row, rotates until it 

 picks out a disengaged trunk, passing over any which are in use 

 by other subscribers. The second movement of the dial operates 

 the second selector in precisely the same way, raising its arm 

 to the sixth row of contacts and causing it to rotate over that 

 row until it picks out a disengaged trunk line leading to the 

 group of subscribers with numbers beginning with 5600. The 

 remaining two movements operate the selector switch and are 



