LXIV ROYAL SOCIETY OF CANADA 
that the above difference between the two observatories of 0’.94 cannot 
at present be regarded as real. 
The horizontal intensities obtained with magnetometer No. 20, were 
on the average too low at Agincourt by 0.000248 c.g.s. unit or 0.00151 H 
and at Cheltenham too low by 0.000456 c.g.s unit or 0.00228 H. The 
correction hitherto assumed by the United States Coast and Geodetic 
Survey for magnetometer No. 20, as the result of various comparisons, 
has been +0.002 H. Whether the difference in the correction to mag- 
netometer No. 20 as disclosed by the comparisons at the two observatories 
is to be attributed to an actual difference between the two sets of 
standards or is due to the possibility that the correction required for 
magnetometer No. 20 may vary with change of magnetic field, cannot 
at present be settled. 
The dip comparisons show that the two observatories agree as well 
as can be expected when it is not possible to make the comparisons with 
and between earth inductors and instead dip circles must be used. 
Corrections obtained for dip needles, depending in most instances upon 
slight irregularities of figure of pivots of the needles, cannot be regarded 
as constant for even comparatively small differences in dip or magnetic 
latitude. 
From a letter received by Dr. Bauer from Mr. O. H. Tittmann, 
Superintendent United States Coast and Geodetic Survey, dated March 
7, 1907, the following extract is made: “ On account of this unsatisfac- 
tory condition of Declinometer No. 26, the results for declination 
furnished you in my letters of November 23rd, and January 31st, will 
need to be increased, probably by nearly one minute. This will cut 
down the difference between Magnetometer No. 20 and the Cheltenham 
standard to a very small quantity, and will bring the standard instru- 
ments of the Agincourt and Cheltenham Observatories into agreement, 
within the error of observation.” 
