1886.] NEW YORK ACADEMY OF SCIENCES. IGl 



graph in the observatory. The beats of the standard clock of 

 the time system were recorded on the chronograph along with 

 our clock-beats and, when star observations were being made, in 

 the middle of the set of observations. 



In examining these tables we note large discrepancies. These 

 discrepancies are too greatand varying to be caused wholly by dif- 

 ferences in the resistances in the electric circuits. Some part of 

 the differences is undoubtedly so caused, but we believe tiie larger 

 part due to careless sending of the signals, and unwarranted as- 

 sumptions as to clock rates. 



In order to get rid of such discrepancies as far as possible, 

 some time ago the following scheme was drafted at Columbia 

 College. An interested friend Jigreed to subscribe the money 

 necessary to put the scheme into operation, and President Bar- 

 nard was heartily in favor of the plan. As yet it has not been 

 possible to secure the co-operation of the surrounding observa- 

 tories : 



A Normal Time Service for Observatories. 



The main features of the plan suggested by Mr. J. Rayner 

 Edmands, of Harvard College Observatory, at the Montreal meet- 

 ing of Am. Assoc. Adv. Science, are: 



1. A normal clock, whose signals are sent daily to each co- 

 operating observatory. 



2. A system of mailing and telegraphing, by which each ob- 

 servatory receives the results of observations made at the others, 

 relative to the errors of the normal signals. 



3. Such work at the central bureau as shall tend to improve 

 and expand the system. 



The complete development of the first feature would involve 

 the erection of a pair of clocks as alternates, with all practicable 

 refinements to secure uniformity of rate. Automatic apparatus 

 for making comparisons might be added. 



To carry out the second feature, each determination of the 

 error of the signals would be mailed directly to such observatories 

 as it would thus reach in time to be useful. It would also be 

 telegraphed to any observatory at which it promised to be par- 

 ticularly valuable, judging from weather reports. The central 

 bureau would receive all results, discuss them systematically, and 

 keep the observatories posted to as late a date as may be. The 

 third feature covers the discussion of systematic discrepancies 

 between the several observatories; experiments to guard against 

 error in transmitting signals; the printing of blank forms for 

 mailing; improvements in apparatus, etc. 



By this plan each observatory has the benefit of an additional 

 normal time-piece, while there is distributed a scientific time- 



