a | 
>) 
1878.] 37 [Frazer. 
existence, yet it must be apparent to any one who has compared them to- 
gether that there are generally discrepancies between them. 
For instance, three authorities which should command the confidence 
of scientific men give the following values : 
Rankine. | Crookes.- | Eliot & Stover. 
Grains inva Grams 2... | 15.48235 | 15.488395 15.4346 
Cubic metersin 1 cubic foot) 0.0283153 | 0.028314 |.............. 
Tonnes ipa ton s..<0. <1: +) 1.01605 | 1.015649 Maes walslauieias 
Kilos. per sq. centimeter in) | 
one pound per sq. inch. .| .0703095 | OU CA Ne etaver let acss ot aronens: 
Only three authorities are here quoted, but the number might be almost 
indefinitely increased. It is true that for most purposes these differences 
being less than one thousandth of one per cent., would not seriously affect 
the results ; but there are problems continually occurring where some rec- 
ognized equivalent is most desirable, and still a greater number where it 
is desirable that all the diverse terms employed should have been obtained 
from the same original unit and by the same methods. 
It would be far better that all the English speaking world should accept 
a wrong determination as the only Jegal one than that each person who 
employs such reciprocal values should take a different standard, even if one 
of the number could be absolutely right. 
In all questions relating to the value of lineal, superficial and cnbical 
equivalents of the English and Metric units, including those defined by 
law as a certain whole number and fraction of cubic inches or feet (e. g. 
the bushel, barrel, stone-perch, &c.), the determination of Kater has been 
taken, and squared, cubed, multiplied and divided until the expression for 
the desired derivative of the meter was obtained in terms of some deriva- 
tive of the inch, no decimals having been omitted until the final number 
was reached ; when the shorter approximative expression has been sub- 
stituted by an application of the well-known rules governing such cases. 
The number of decimal places given has been in proportion to the im- 
portance of the unit as a base from which to calculate other values. Thus 
the number of places in the Grain-Gram equivalents is eleven (as in the 
report of Mr. Upton, from which it was taken), whilst the Rood—Are being 
less frequently used and especially being of less importance as a base from 
which to derive other values, is given in five and six places respectively. 
This method of separate calculation from the fundamental Inch-Meter 
value has been employed for each of the above-mentioned kinds of dimen- 
sions, and the value of the metric unit in the Inch derivative has been con- 
verted into the reciprocal or Inch derivative unit by simply dividing the 
whole decimal into one and shortening as before. This is obviously to be 
preferred to taking the reciprocal of the legal value of the meter in 
inches, as the base of the calculation. 
Crookes’ (Select Method of Chemical Analysis) was drawn on for the 
PROC. AMER. PHILOS. soc. xvilI. 101. 80. PRINTED MAY 18, 1878. 
