448 Experiments to determine variations in length 
to the westward of the Connecticut over the Farmington valley. 
But it follows from the facts presented that the diminution 
outh of Farmington the coarse deposits may have been 
made with the slope diminished a foot or more. : 
In the Meriden valley, where the terrace formation consists 
mainly of sand, the present pitch of the terrace-level is much 
too great for such depositions. From Hanover Pond to North 
Haven the mean slope, according to Table II, is 44 feet a mile; 
and about 4 feet, from the same point to Fair Haven. 
Art. LXVIIL—Results of some experiments made to determoné 
the variations in length of certain bars at the temperature 
melting ice; by R. S. Woopwarp, KE. S. Waeeer, A. 
Fuiint and W. Voter. 
THE precision attained of late in comparisons of standards of 
length, and in geodetic work dependent on such standards, has 
rendered the question whether a given bar can have differing 
lengths at the same temperature an important one. In ord 
to obtain some data bearing directly on this question the au 
thors of this paper have undertaken on their own account? 
series of experiments with bars of various metals. : 
For the purpose of making these experiments the following 
apparatus has been provided : 
Ist. I'wo micrometer-microseopes designated F and W a’ es 
spectively. The optical work on these is by Bausch and eee : 
of Rochester, N. Y., and the micrometers and stands were mat’ — 
