C. G, Rockwood—Motion of a Tower by Solar Heat. 181 
somewhat the extreme ordinates; as we see that by simply 
lengthening the hee i 
ordinates on the Psi ee: oN 
Dene BRE er SIN et } 
axis be brought / : = 
into the merid- { i 4 
ian. 2 ee 
. . ‘ 2 pee 2 
But it is evi- seo cee 1 a ae 
y: 
of observation, were as follow: 
Mean Level Readings—Clear Days. 
+East, “2 50ma.M. 10h88ma.m. Oh59mP.M. 8h 32m P.M. 558m P.M. 10h 39m P.M 
ot ea ee ee Se 
46/520 garages = — 855 — 886 tT EL 
Mean Level Readings— Cloudy Days. 
r M. Oh59mP.M. Sh33mP.M. 6hSmP.™M. 10h 47m P. M. 
T  —9'-94 geo 8/65 = —10"15 11/43 11798 
TL +592 4676 4.559 + ANA + BOT + BI GS 
These averages, with the signs changed, being made erage od 
om ordinates, give the two curves oo B and OC, the obser- 
ons being indi numerals as before. 
. 2 epee result was the 
Separation into two series. The two curves have indeed = 
_, ls excessive departur : “14: 
attributed solely 0th connection with the main building, for 
