358 
COLONEL SABINE ON PERIODICAL LAWS DISCOVERABLE 
ences of the scale-readings in different natural temperatures, when the magnetometers 
were mounted and in use. 
The formation of the Monthly tables of the “ scale-readings reduced to a uniform 
temperature of 55°,’’ from the tables uncorrected for temperature,” was performed 
by two Non-commissioned Officers of the Royal Artillery, working independently of 
each other ; the correctness of their work was proved by the accordance of the two 
independent computers; the daily and hourly means were then taken in the same 
manner by two independent computers, and were additionally checked by their com- 
parisons with the daily and hourly uncorrected means calculated at Toronto; these 
means being also reduced, for the purpose of this comparison, to the standard tempera- 
ture of 55° (excepting in a very few instances, in which the observations on days of 
excessive disturbance had been omitted in the sums and means of the uncorrected 
readings computed at Toronto, but were restored in the sums and means of the cor- 
rected tables). The new tables thus formed, of the scale-readings reduced to 55°, 
then passed into my hands, when I satisfied myself, by a careful examination, 
that a difference of fourteen scale-divisions for the Bifilar magnetometer, and of four 
scale-divisions for the Vertical Force magnetometer, above or below what might be 
taken as a normal value, — viz. the mean value at the same hour during the same 
month, or for several preceding and several succeeding days, — would constitute a con- 
venient minimum limit for the disturbances of largest amount; since, on the one 
hand, it would be a greater departure from the normal value than could reasonably be 
ascribed to any other cause than that of a disturbance in the earth’s magnetism, whilst, 
on the other hand, the number of disturbances that would be thereby separated, would 
form a sufficient body to permit their periodical laws (if such existed) to be investi- 
gated. Having determined on this limit, I proceeded to mark provisionally, with a 
pencil, every observation which differed in the Bifilar fourteen scale-divisions or more, 
and in the Vertical Force magnetometer four scale-divisions or more, from their respect- 
ive normals. I then recomputed the normals, omitting the observations provisionally 
marked as disturbed, and compared afresh all the observations, including the provi- 
sionally-marked ones, with the new normals, altering the markings where required ; 
and I continued this process until the normal in every case included every observation 
which differed in the Bifilar less than fourteen, and in the Vertical Force magneto- 
meter less than four scale-divisions from itself, and excluded every observation which 
differed in the Bifilar fourteen, and in the Vertical Force magnetometer four scale- 
divisions or more from itself. The excluded observations were then marked, finally, 
with a surrounding ring in ink. In this state the Tables were returned to the Office, 
and the correctness of the markings, and of the normals excluding the larger disturb- 
ances, was examined by a separate computer. 
Two computers, working separately, and having their work compared, then formed 
a table of the marked disturbances of each element, during the five years, arranged 
