256 
DR. C. CHREE: DISCUSSION OF KEW MAGNETIC DATA 
Daily Range. 
§ 25. The term daily or diurnal range is used in several senses. It may mean the 
difference R between the algebraically largest and least hourly values in the diurnal 
inequality, that inequality being derived from selected days or from all days. It may 
mean, however, the difference It' between the highest and lowest daily values, 
irrespective of the time at which they occur, whether an exact hour or not. The 
mean It' for a month is simply the arithmetic mean of the values for individual days 
of the month. 
Tables XXXVI. and XXXVII. give It for H and V from ordinary days, for each 
month of the eleven years. Tables XXXVIII. and XXXIX. give It' from all days for 
the same two elements. The It derived from any combination of days must be less 
than the corresponding It', unless the maximum and minimum each occur at a fixed 
time, which is an exact hour. In practice, the times of the maximum and minimum 
vary from day to day. Speaking generally, the difference between It and It' for a 
particular month is larger the more disturbed the month. 
If instead of the month one takes the year, the mean It' is the arithmetic mean of 
the values for the 12 months, but It is less than the arithmetic mean of the monthly 
values unless the hours of maximum and minimum in the inequality are the same 
throughout the year, which is never the case at Kew. Even when one considers the 
months of the same name from a number of years, there is usually some variation in 
the hour of maximum or minimum. To bring out these points, Tables XXXVI. and 
XXXVII. contain the values of II from the inequalities for the whole year and for all 
the months of the same name combined, as well as the arithmetic means of the values 
for the individual months. 
In Table XXXVI., November is the only month in which the R from the months 
combined equals the arithmetic mean of the values for the separate months. In the 
other 11 months the former quantity is the smaller, though the difference is never 
large. There is a much larger difference between the inequality range for the whole 
year and the arithmetic mean of the values of R for the twelve months. The former 
quantity on the average stands to the latter approximately in the ratio 7 • 8. 
In Table XXXVII., April is the only month in which R from all the months 
combined is the same as the arithmetic mean of the values for the individual months, 
but the differences in the other months are small, just as in the case of IT. The 
differences in Table XXXVII. between the inequality range for the year and the 
arithmetic mean of the ranges for the twelve months are a good deal smaller than in 
the case of H. The average excess of the arithmetic means is only about 3|- per cent. 
This implies, of course, less variability in the hours of maximum and minimum. 
Comparing Tables XXXVIII. and XXXIX. with Tables XXXVI. and XXXVII. we 
see that the excess of R/ over R is usually large. Allowance must, of course, be made 
for the fact that the more highly disturbed days—242 for H and 348 for V— 
