TEE TRANSATLANTIC LONGITUDE. 15 



sets produced by the former are of practically equal length, this length dependmg 

 on the adjustment of the armature and strength of the battery ; while those produced 

 by the observation-signals are for a -practised observer quite near enough to equality 

 to preclude any difficulty in reading off the records, except in very rare instances. 

 For portable instruments there seems to be no room for reasonable doubt as to 

 the superiority of an instrument with a single pen ; and for the fixed instruments 

 of an observatory I should personally give this construction a decided preference. 

 All signals are given by the interruption of a closed circuit, so that, when the 

 observing key is properly adjusted, no interval elapses between the first pressure 

 and the transmission of the telegraphic signal ; while the moment of release of the 

 armature from the electro-magnet is distinctly recorded. The clocks are all provided, 

 according to Saxton's plan, with delicate platinum tilt-hammers resting on platinum 

 disks, and so adjusted that a small pin fixed in the pendulum-rod at its centre of 

 percussion shall strike the tilt-hammer at the instant when the rod is vertical, and 

 thus lift the hammer from the disk for a very brief period, generally about the one- 

 hundredth part of a second. The galvanic circuit to the chronograph being con- 

 ducted through this tilt-hammer and disk, the circuit becomes interrupted for a 

 moment at each oscillation of the pendulum. 



The advantages of this mode of recording the clock-signals over any in which 

 the galvanic current traverses any portion of the clock itself, or in which the sig- 

 nals are produced according to Saxton's original plan by contact with a globule of 

 mercury, have been sufficiently set forth in previous reports, and require no repeti- 

 tion here. 



IV. 



OBSERVATIONS AT VALENCIA. 



Here the Krille clock and Transit-instrument No. 4 were employed. I had sup- 

 posed all precautions taken to icsure that the instruments should be in good order ; 

 but, owing probably in part to the haste with which the expedition was organized 

 in view of the approach of winter, this was not the case, and the want of proper 

 condition of both these instruments, as well as of the minor telegraphic apparatus, 

 much augmented the unavoidably serious difficulties of the enterprise. 



Observations were obtained on fifteen nights during our sojourn at Valencia, on 

 no one of which the sky was unclouded. On only two of the five nights on which 

 longitude-signals were exchanged with Newfoundland was it possible to obtain 

 observations after the exchange, and this was possible, too, on only one of the three 

 nights when signals were successfully exchanged with Greenwich. Observations 

 of circumpolar stars for the special purpose of determining the intervals of the transit 

 threads, were out of the question. Indeed there was but one instance when a transit 

 of any star north of 60° declination was observed over all twenty-five threads. In 

 those rare instances when this would have been possible, the stars were needed for 

 determining the error of coUimation. 



At the close of the series of observations, it was found that 53 complete transits 

 had been observed over all the threads ; and since the equatorial intervals of the 



