126 
PROFESSOR HUGH L. CALLENDAR ON 
The standard value of Js 0 assumed in reducing the tables in the case of water for 
calculating the value of D, was the British Association unit of 4‘200 joules. Inserting 
this value, and jmtting a — '0039, we have the numerical formulae 
Correction to h = + ‘00039 hdO — T05/i 2 /Q' — T05/i 3 /Q r/ . 
Correction to J = + T05/i 0 3 /Q'Q".(15). 
These formulae are to be employed in correcting the results for the specific heat of 
water given in the summarized tables (Barnes, p. 243). 
In cases where more than two flows are available at the same time and under the 
same conditions, it is naturally possible to obtain a more reliable value for the result 
by utilising all the Hows, so as to minimize the effect of accidental errors. The 
accuracy of the work may then be verified by comparing the observed values of D for 
each flow with those calculated by equation (12). As an illustration we may take 
observations I. and II. with calorimeter D (Barnes, p. 243), which are also selected 
by Dr. Barnes as an experimental verification of the elementary theory. 
Table IX.—Correction for Variation of Gradient in Flow-Tube. 
Specific Heat of Water. Observations I. and II. (Barnes, Table XVIII., p. 243), 
Calorimeter, D. 
Number of flow 
(1) 
C) 
(3) 
(4) 
(5) 
Mean of 2,3,4 
Flow, Q, gramme per second 
•6741 
•3993 
•3902 
•4967 
•2482 
•4287 
ho, watts per degree 
•06975 
•06975 
■06975 
■06975 
•06975 
•06975 
■0039h 0 cW/l0 .... 
- -00021 
- -00022 
- -00022 
- -00022 
- -00022 
- -00022 
• 105/n 2 /Q. 
+ -00076 
+-00128 
+ -00132 
+ -00103 
+ -00207 
+-00120 
. 
Oi 
O 
r-H 
o 
1 
- -01281 
- -00758 
- -00741 
- -00944 
- -00471 
- -00815 
D, calculated .... 
(-05749)* 
•06323 
•06344 
•06112 
•06689 
(•06258)* 
D, observed .. 
•05749 
•06317 
•06318 
•06135 
■06691 
•06257 
■ 07145--00207Q . . . 
(-05749)* 
•06318 
•06337 
•06117 
•06631 
(•06257)* 
The last column gives the mean of the three intermediate flows, which has been 
combined with the first flow in the calculation of the results. The second line gives 
the approximate values of the flows, which are required in working out the results. 
The next four lines give the values of the several terms in equation (12) for the 
calculation of D. The two unknown quantities h 0 and J sd are calculated by assuming 
the values of D enclosed in brackets for (l) and the mean of (2, 3, and 4). 
The values of D “ calculated ” are obtained by adding the numbers in the four 
lines above, and are corrected for the effect of variation of the gradient in the fine 
flow-tube. The values D “observed,” given in the next line, are taken from the 
tables, and are found by subtracting 42Q from EC/dO. 
* Assumed in the calculation of h 0 and d. 
