go Captain Kater's account of the 



the cone is turned very smooth and close to its base is curved 

 so as to avoid the acute angle, which would have resulted 

 from continuing the side of the cone to the bottom. By this 

 also the advantage was gained of a greater substance for the 

 insertion of the screws used for attaching the bottom of the 

 gallon to the cone. 



As soldering was thought objectionable, the bottom was 

 ground to the cone, and it was supposed that it would thus 

 have been sufficiently secure ; but on letting the gallons 

 remain filled for 24 hours, in some a slight leakage became 

 perceptible. After many experiments, the best remedy for 

 this appeared to be a very minute quantity of fresh grease, 

 carefully applied to the bottom of the cone and then almost 

 wholly wiped off; after which the bottom was firmly screwed 

 in its place. 



The following was the method pursued in adjusting the 

 gallon : 



If the gallon contained a considerable quantity, as one or 

 two hundred grains of water too little, its capacity was 

 enlarged by turning away a small portion from the flat 

 bottom. If on the contrary the gallon contained too much, 

 the base of the cone was ground away ; and it must be evi- 

 dent that a very small quantity taken from this part would 

 occasion a considerable variation in the capacity, one thou- 

 sandth of an inch making a difference of about 17 grains. 

 Errors of smaller magnitude were removed either by grind- 

 ing down the top of the mouth, or by enlarging the aperture 

 by using a cylindrical plug of brass with emery. 



The interior of the measure being carefully wiped dry, 

 it was placed in the scale together with one of the gallon 



