34 



SCIENCE. 



[Vol. II., No. 23. 



tions of certain of the other levers in the 

 frame. To exemplifj' this, we will take three 

 levers, A, B, and C. If A and B be in such 

 position that a signal given b}' the movement 

 of lever C will be dangerous or misleading to 

 a train, the pivoted bar connected to lever C 

 is locked, and cannot be moved b}- anj^ exer- 

 tion of strength on the part of tlie signalman ; 

 and therefore he cannot even begin to move 

 lever C, and the possibilitj- of giving a wrong 

 signal is put bej^ond doubt. Similarly, nothing- 

 is effected unless the lever completes its stroke. 

 The pivoted bar or ' rocker,' through which the 

 whole work of interlocking is done, moves 

 only at tlie extreme ends of the stroke of the 

 levers, and then is only moved b}- the rising 

 or falling of the spring detent. This inven- 

 tion, simple as it seems, is the result of many 

 ^•ears' experience, accidents having often oc- 

 curred through a lazy signalman pulling his 

 lever through part only of the stroke, and thus 

 onh- partiallj' effecting the locking. This is 

 now impossible ; and the intention of a switch- 

 man to move a lever, expressed hy his grasping 

 the lever and so moving the spring-catch, inde- 

 pendeutlj- of his putting the intention into 

 force, actuates all the necessary locking. 



The details of locking-apparatus are some- 

 what complicated, but the principle is simple. 

 Certain bars carrjing lugs or projections are 

 made to slide or move by the movements of the 

 rockers. Certain other bars, which are also 

 moved bj- the action of one or more rockers, 

 are slotted or pierced with holes, so that, in 

 certain positions, the lugs in the first set of bars 

 can enter the holes in the second set of bars, 

 and, in other positions, the lugs strike against 

 the bars, and cannot be moved. It is, of 

 course, obvious that the arrangement is such 

 as to prevent unsafe or contradictor^' signals 

 being given, and permit onh' of safe or harmo- 

 nious signals; and, bj' a careful arrangement 

 of the locking-appai'atus, it is sometimes pos- 

 sible to make a few movements effect important 

 changes of the switches and signals with a 

 minimum of levers and complication. 



It is obvious, that, when switches are worked 

 from a distance, there is a chance of the switch 

 being incompletely' closed, owing either to 

 dirt, or a stone, or ice, choking the switch 

 itself, or the switch-rods working it. There 

 is also a danger that the switch-rod might 

 break or become disconnected, and that, 

 though the signalman moved all his levers, and 

 all the locking and unlocking was properly per- 

 formed in his cabin, j'et the switch itself might 

 remain unshifted, or be left half open. To 

 obviate this, the facing point lock was invented. 



This is a bolt which can onl3- be shot into 

 a crossbar connecting the two rails of the 

 switch when the switch is either properly 

 closed, or wide open. A failure of the switch 

 connections, or an obstruction in the switch, 

 will render it impossible for the bolt to enter 

 the opening to lock the switch ; and, as the 

 signalman's lever actuating this lock inter- 

 locks with the signal levers, no train can be 

 signalled to approach until the switch is either 

 closed, or wide open, as the case may be, and 

 firmly locked in its proper position. But an- 

 other danger has to be guarded against : signal- 

 men, to save time, will generally throw a signal 

 again to danger directly the engine of an ap- 

 proaching train has passed ; his other levers are 

 then set free, and he can unlock his switch, 

 and actually change the switch, before the whole 

 train has passed, thus probably throwing the 

 rear vehicles off the track, and causing a 

 serious accident. To guard against this, a 

 locking or detector bar is used, which lies near 

 the rail, but clear of a wheel, when the switch 

 is either shut or full open ; but directlj' the 

 switch is moved from either of these positions, 

 the bar moves close to the tread of the rail, and 

 takes such a position that it must come in 

 contact with anj* wheel approaching the switch. 

 As the bar is made longer than the distance 

 between anj' two trucks, it follows, that, as 

 long as a train is passing over the switch, one 

 or more wheels of the train must prevent this 

 bar being moved, and, as the switch-lock and 

 the bar are arranged to move together, it fol- 

 lows that the switch cannot be unlocked until 

 the last truck of the last car of a train has 

 passed. The Union switch and signal com- 

 pany adheres to Saxby and Farmer's arrange- 

 ment of this bar where it moves verticallj'. 

 The Pennsylvania steel company shifts it later- 

 ally. The latter movement is more easily per- 

 formed, and the bar can serve as a guard-rail ; 

 but its movement seems somewhat liable to be 

 impeded bj' snow falling between the rail and 

 bar. 



{To be continued.) 



THE WEATHER IN MAY, 1883. 

 There have been two periods of ver^- severe 

 storms, and at manj' places of tornadoes. The 

 first of these accompanied a ' low,' first noted 

 in Colorado ^ on the 13th. This moved with 

 considerable energj' over Colorado and Ne- 

 braska. On the 14th, increasing in energy, 



1 It has been found necessary, owing to the smallneee of the 

 appropriation, to give up all telegraphing reports west of the 

 Eocky Mountains : hence the charts arc made up only to the east. 



