70 BULLETIN OF THE NATURAL HISTORY SOCIETY. 



end rested on a spring by means of a similar axial rod having a 

 conical point which turned with ver}' little friction in a conical 

 hole in a brass plug attached to the spring. The purpose of the 

 spring was to keep the cylinder pressed tightly against the clock. 

 A long springy piece of brass was soldered by its lower end to- 

 the side of the float and its upper end carried a pencil which 

 pressed lightly against the cylinder. It is easily seen that sur- 

 face waves might move the float and so obscure the tidal record ; 

 hence the whole instrument was enclosed in a long, narrow, 

 vertical box which leaked slightly at the bottom. Thus the- 

 water level in the box changed with the slow rise and fall of the 

 tide but surface waves had no effect. The machine was held at 

 the pi'oper level in the water b}' being solidly clamped to an iron- 

 stake driven in the ground. 



In preparing for an ol)ser\ ation, a sheet of white paper vvas- 

 wrapped around the cylinder. The cylinder was then put ini 

 place and the pencil arranged so as to press against it. The 

 exact time of beginning and ending the i-ecord being noted, the 

 time corresponding to any particular point on the curve could be- 

 deduced after the paper was removed. 



VI. Curves Obtained at 8prixgiiill. 



The first point at which this instrument was used was imme- 

 diately aliove the rapids aV)Ove Springhill on July 21st and 22nd. 

 The line traced in twenty-four hours indicated a fall of thirteen- 

 sixteenths of an inch, but diflfered by less than one-sixteenth of an 

 inch fi'om a straight line, indicating that absolutely no tides are- 

 propagated above these rapids, at least along the right bank. The 

 next point chosen was just below the rapids, about a (juartei- of 

 a mile above the Springhill hotel. A twentj'-seven hour record 

 was taken on July 2.'lrd and 2ltli. This showed in a remark- 

 able way an effect frecjuently afterwards noticed, the great influ- 

 ence of wind. From Cy.l'tO p. m. to 9 p. m. of the 2."b-d, a strong 

 wind hU'W down stream, and duiing this time the pencil traced 

 almost a straight line. At 10. .'JO it rose about a (piarter of an 

 inch, and then fell smoothly to low water at 3 a. m., and rose to- 



