La 
OBSERVATIONS AT WINNIPEG. 91 
the Lord Bishop of Rupert’s Land—for the erection of a shelter, which with some 
slight additions proved suitable for the purposes of our work. The shelter was ten feet 
by fourteen and was divided into two compartments. The roof which was flat and sloped 
towards the south was made—in part movable in the telescope compartment, giving an 
uninterrupted view of the sun during the time of the transit. The transit room was pro- 
vided with the usual opening. 
In passing through Chicago I was indebted to the kindness of Colonel Clowry, man- 
ager of the Western Union Telegraph Company, who granted the free use of the company’s 
lines for the purpose of transmitting time between Chicago and Winnipeg. 
I have also pleasure in recording the kind attention of Professor Hough of the Dear- 
born observatory. 
Through the courtesy of the Great North Western Telegraph Company and the mana- 
gers of the Winnipeg office the line from Winnipeg to Selkirk was cut and passed 
through our observing station. This gave connection with the Dearborn observatory at 
Chicago through relays at St. Paul and Chicago. 
Time Signals for Longitude. 
The signals received were those of the clock at the Dearborn observatory, Chicago. 
This clock closes the circuit at intervals of two seconds, up to the 52nd second for each 
minute except before every fifth minute when the last contact is made at the fortieth 
second. 
December 4th, P. M. Signals were received during 12 minutes; observed differenee 
in time 0" 38" 02:95. 
December 5th, P. M. Signals were received during 3 minutes; observed difference 
in time 0° 38" 02:3. 
Mean of differences on December 4th and 5th 0° 38" 02:62. 
Allowance for armature and current time 00°12. 
Reduced difference in time 0" 38" 02:74. 
Longitude of Dearborn observatory 5" 50" 26°78. 
Longitude of Winnipeg station 6°28" 29°52. 
Time signals were also received and sent on the night of December 8th, but as no 
time observation was made subsequent to the 7th and the rate of the standard chrono- 
meter was not very well determined, I do not consider that a reliable difference of time 
can be obtained from this exchange. This was the only night on which my signals were 
received at the Dearborn observatory. My signals were sent by hand and received on a 
chronograph. The difference of time thus recorded as given to me by Professor Hough is 
precisely the same, to a tenth of a second, as the difference observed by me. 
These times should differ by twice the armature and current time. This error is no 
doubt almost entirely due to my imperfect sending which besides being by hand was 
otherwise not under favourable circumstances. The signals received could be compared 
‘with the chronometer to the nearest tenth of a second with ease. On account of these 
circumstances I have thought that greater accuracy would be obtained by making the 
determination of longitude depend, on differences obtained in one direction only, adding 
an armature time as above. 
I consider the above result to be within one second of the truth. 
