314 PROCEEDINGS OF THE AMERICAN ACADEMY 



centre of the bottom must be attended with the elevation of 

 the whole body of the can, and the sudden downward move- 

 ment of the bottom would cause the can to spring into the 

 air. The shelf was inclined toward the sink, and the outer 

 half inclined also a little outward. This inclination would 

 give the upward movement of the can a direction from the 

 perpendicular, and, if the can were seated on the outer half, 

 an inclination outward from the shelf and sink. The latter 

 supposition is a little more favorable to the view taken, but not 

 essential. With a velocity that would carry the can to the 

 inclined roof, it is easy to see how the nose could have been 

 broken (the neck and plug having been separated by the ex- 

 plosion), and, with the momentum acquired, how a quantity of 

 fluid would rush out upon the rafter or inside of the roof, and 

 some of it fall. The can, as the resultant of the collision of 

 its irregular form with the inclined inner surface of the roof, 

 would acquire more or less of a whirling motion, and, scatter- 

 ing fluid in its way, would ultimately reach the floor. A jet 

 of it falling upon the stove would instantly enshroud -it and 

 the girl by its side in flames. The heat of the burning fluid 

 about the can would melt the solder, release the bottom, and 

 such portions of the soldered seams as were not protected by 

 the fluid. The line of attachment of the conical top to the 

 sides, the opened seam of the top itself, the undisturbed ear, 

 to which the pail was on one side secured, and the gathering 

 of the molten solder in the same region, all are in keeping 

 with the idea, that the can lay partially immersed, and so far 

 protected, by the fluid on the floor. To return to the point 

 of collision of the can with the roof. What point on the 

 shelf would a small quantity of fluid reach, thrown from the 

 neck of the can at the instant of its collision with the roof, 

 and falling perpendicularly ? A point manifestly lower on the 

 inclined shelf than that occupied by the can ; and although it 

 may not now be susceptible of absolute demonstration by ad- 

 measurement, since the exact position of the article is not 

 known, it is obvious, upon an inspection of the premises, that 



