348 Messrs. J. C. McConnel and D. A. Kidd. [June 21, 



optic axes were nearly parallel to the length of the columns. That 

 the axes, however, were not accurately perpendicular to the length 

 of the columns, i.e., horizontal in the original position on the lake, 

 was shown by examining separate columns. After allowing the 

 ice to thaw slightly, or better after leaving it in the rays of the sun 

 for twenty minutes, the columns could be easily separated. 



Detailed Account of the Experiments. 



It will be more convenient to describe all the experiments made 

 with Main's apparatus first, than to keep to the chronological order. 



Experiment No. 1. Mairi's Apparatus. — Measurements were taken 

 with the cathetometer. The specimen was a square bar of ice, taken 

 from the surface of a bath of water about a foot deep, and cut into 

 shape with a knife. It was perfectly clear and free from bubbles. 

 It was wrapped in gutta-percha tissue, which was not removed till 

 the end of the next experiment. The " needles " were pieces of steel 

 knitting needle. The area of the section was 8*1 sq. cm., and the 

 tension 3*7 kilos, per sq. cm. 











Difference between the 





Distance 





Tempera- 

 ture. 



two sides. 



Date. 



between 



Extension. 









needles. 





Upper. 



Lower. 





mm. 



mm. 





ram. 



mm. 



Jan. 14, 11 h 



163 93 







-3*0° 



4-1 



4-6 



„ 16, 9h 



164 -06 



+ 0-13 



-8-0 



4-5 



4 3 



17, .... 



163-91 



-0-02 



-7-0 



43 



4-2 



,. 18, „ .... 



164 -04 



+ 0-11 



-62 



4-3 



4 2 



„ 19, „ .... 



163-98 



+ 0-05 



-55 



4-3 



4-2 



„ 20, „ .... 



164-08 



+ 0-15 



-5-0 







,, >, .... 



164 -01 



+ 0-08 









„ 21, „ .... 



164-13 



+ 0-20 



-4-0 



4 4 



4-i 



The hours in the first column are reckoned from midnight. The 

 third column gives the extension observed, measured from the length 

 at the first reading. The fourth column gives the temperature just 

 before each reading. The maximum temperature during the whole 

 period was — 3"0° and the minimum — 8'5°. The fifth column gives 

 roughly the difference between the heights of the marks on the right 

 and left ends of the upper needle, and the sixth column the same 

 thing for the lower needle. These are added to show that a slight 

 bending took place chiefly between the 14th and 16th. On removing 

 the gutta-percha at the end of the next experiment, a surface crack 

 was found which may have occurred at the same time. Each reading 



