39 8 Mr. Macdonald's Observations of the 
as philosophy may draw ; and tends to confirm some inferences 
stated in a former Paper, containing similar observations taken 
in the East Indies. By adding the mean of the morning and. 
afternoon observations, at St Helena, and taking the half, the 
general variation, in the month of November, 179b, appears to 
have been iy° 48' 34/- west : and, by subtracting the medium 
diurnal afternoon variation, from the medium diurnal morn- 
ing, the vibrating variation proves to be f 55". It appears, that 
the magnetic needle is stationary from about six o'clock in the 
evening till six o clock in the morning ; when it commences 
moving, and the west variation increases, till it amounts to its 
maximum, about eight o'clock ; diminishing afterwards, till it 
becomes stationary. Here, the same cause seems to operate as 
at Bencoolen, with a modification of effect, proportioned to the 
relative situations of the southern magnetic poles, and the places 
of observation. At the apartments of the Royal Society, this 
species of variation is found to increase, from seven o'clock in 
the morning till two o'clock in the afternoon. If the variation 
is east, in the northern hemisphere in the East Indies, I con- 
ceive that the diurnal variation will increase towards the after- 
noon, remain some time stationary, and diminish before the 
succeeding morning : if the general variation is west, in that 
quarter, the reverse may be the case. The quantity of the diur- 
nal variation is greater in Britain than at St. Helena, or at Ben- 
coolen. This will naturally arise from this country's being 
more contiguous to its affecting poles, than those islands si- 
tuated near the equator. It were to be wished, that observations 
were taken in as many situations as possible, similarly situated 
in the opposite hemispheres, on the lines of no variation. A 
greater degree of dip might be found, and conclusions might 
