C. G. Rockwood—Motion of a Tower by Solar Heat. 179 
A third level placed with the others upon the stone floor, 
and also the levels of a zenith telescope, which was temporarily 
mounted on a brick pier in the room, were recorded during a 
part of this time, but as their results were not employed in the 
discussion, it is not necessary to notice them farther than to 
say, that in general they confirmed the indications of the prin- 
cipal levels. 
et us now examine briefly what would be the probable 
motion of a tower thus situated, and then compare this theoret- 
leal result with that given by the recorded level readings 
First : suppose an isolated symmetrical tower of homogene- 
ous material, situated at the equator of the earth, and the sun 
at the equinox. The diurnal circle of the sun then passes 
through the east and west points of the horizon and the zenith 
of the tower. In the morning the heat of the sun’s rays would 
expand the east side of the tower, and cause it to lean toward 
the west. As the sun rose toward the zenith, and warmed 
Rormal position. 
t any station intermediate between the equator and the 
Pole, the figure described by the tower would be neither a 
ht line nor a circle, but between the two, i. e, an ellipse, 
Whose excentricity diminishes as the latitude of the place 
creases, 
In the case under discussion, the tower is in N. lat. 41° 19’, 
4nd the sun was near the summer solstice. The sun therefore 
Tose about 80° north of the east point, culminated south of 
“€ zenith, and set north of west. 
The inclination of the tower, being opposite the sun, would 
.-1 1 the morning southwest and west, at noon north, at even- 
‘ng east and southeast, and during the night it might be sup- 
to return in a straight line to its mean position. Since 
Melination would be less than the westward or eastward, and 
_ the curye described would resemble an ellipse, with its wept 
Pat the meridian, and probably somewhat flattened on the 
