396 
PROF. LOUIS YESSOT KINO ON THE CONVECTION OF 
At the junction we have 
H 1 = H 2 +H', 
(55) 
assuming perfect thermal contact between the anemometer-wire and the potential 
lead. • From (55) we are now able to calculate H 7 and therefore from (51) the 
temperature of the anemometer-wire at any point; we find that 
0—0 O = (1 — Q cosh px) . Ffto/Kcop 2 ,.(56) 
where, writing P = ^/(Kco'A^/Kwp, Q is given by the equation 
Q = 
__1 + P sinh pl 0 _ 
cosh p (l +l 0 ) + P sinh pi 0 cosh pi ’ 
(57) 
Let 6 and R represent the temperature and resistance per unit length of an 
infinitely long wire under the same conditions of wind velocity and heating by 
electric current. We then find the relations 
0—0 O = rPfi/Kwp 2 ; p 2 — rLfi/Kw (0 — 0 O ) — If,A/RKo> ; P 2 = a/PA / /«Pt IJ A. (58) 
The resistance R/ measured between potential leads is given by 
ri ri 
R/ = j R 0 [l +a (0—0 O )] dx — R 0 / + R o d (0—0„) I (l + Q cosh px) dx, 
Jo Jo 
giving on integration and making use of the relation R = R„ [l +a (0 —0 O )] the final 
result 
R/R = [l _(1 _R JR) Q (sinh pi) /pi] .(59) 
This result enables us to estimate the effect of the potential leads and terminals on 
the measure of resistance. In the experiments under consideration the quantity 
measured is R ; it is necessary in the determination of the convection constants to 
employ wires of such a length that R may not differ from R by more than a fraction 
of 1 per cent. 
Since A'/A = (/TVWy')/(/3W + y) we may write as a rough approximation since 
from the experimental results the terms y and are small in comparison with the 
first terms of the expressions in which they occur, A'/A = (3'/(3. It also follows 
from the experimental results that ft 1 /ft = (a 7 jaf so that since w'/w = «' 2 /« 2 we may 
write for purposes of rough calculation 
p - (R/Ro) 1/2 («74 5M .(60) 
Writing from (46) A = W/(0 — 0„) we have for purposes of calculation 
p = [R 0 W/EKw (0 —0 o )] 1/2 .(61) 
The quantities P and p are easily determined from the constants given in Table II. 
In the description of Diagram III. are set out the values of these quantities for 
