228 TENTH ANNUAL REPORT OF 



sandtlis of an inch. The divisions of the smaller graduated glass 

 cylinder will measure directly the thousandths of an inch, and it may 

 serve, in case of accident, as a substitute for the larger one. The two 

 glass vessels may be placed in the barrel itself, if it is of sufficient 

 size. They must be placed in a reversed position, on two upright 

 pegs, to let them drip out. As soon as the observation is made, it 

 should be noted in pencil, not trusted to the memory ; and written in 

 the journal upon entering the house. 



SNOW-GAGE. 



Ohservation. — The snoio-gage sh,ould be supported vertically, in 

 an open place, between three short wooden posts, its opening being 

 about two feet from the ground. It should be employed in the fol- 

 lowing manner : 



When only a very small quantity of snow falls, or of snow alterna- 

 ting with rain, or of dry and fine snow, driven by the wind, it should 

 be collected in the snow-gage, as would be done in the ombrometer. 

 But when the snow falls in a sufficient quantity to cover the ground 

 more than an inch deep, the vessel must be emptied, and plunged, 

 mouth downwards, into the snow, until the rim reaches the bottom. 

 A plate of tinned iron, or a small board, may then be passed between 

 the ground and the mouth of the gage, and the whole reversed. In 

 this way a cylinder of snow, of which the base is superficially one 

 hundred inches, will be cut out, and received into the vessel. The 

 oi)eration may be facilitated by placing on the ground a platform of 

 strong board or plank, two or three feet square, on which the snow is 

 received. 



The place selected for this purpose must be one where the snow has 

 not been heaped up, or swept ' away by the wind, and where it pre- 

 sents, as near as possible, the mean depth of the layer that has fallen. 

 In order to take only the snow which may fall in the interval between 

 two observations, the board should be swept after each measurement, 

 and the place designated by stakes. 



Beading. — In the reading of the graduated vessels, the general sur- 

 face of the liquid must be considered as the true height, and not the 

 edges, w^hich are always raised along the walls of the vessel by capil- 

 lary attraction. 



The collected snow must be melted by placing the gage, covered 

 with a board, to prevent evaporation, in a warm room ; and the 

 quantity of water produced measured by pouring it into the glass 

 cylinder. It need hardly be said, that if rain and snow fall the same 

 day, no account will be taken except of what the snow-gage receives, 

 unless the ombrometer has been observed separately after the rain, 

 and the snow-gage after the snow. Care must be taken, in these 

 cases, not to count twice the same quantity of fallen water. 



The rain-water and melted snow-water must be separately entered 

 in the journal in the columns reserved for each. 



During abundant rain-falls, it is well to measure the water more 

 than once a day, or at least immediately after the rain ; and the 



