VOL. XXXVIl.] PHILOSOPHICAL TRANSACTIONS. 547 



is SO made, that its discharge (which is continually decreasing, as the surface of 

 the water subsides in the cavern) shall at any time be equal to the feeding- 

 stream entering at o, in such a case, the syphon must continually run, and yet 

 not empty the cavern, till the feeding-stream at o is sufficiently diminished. 

 But, when the diameter of the syphon at n, according to the height of the 

 cavern, is so great, and the feeding-stream at o so small, that the syphon can 

 carry ofF, in the manner of a waste-pipe, all the water which comes in, and yet 

 not run with a full stream ; the syphon must then continue to run without 

 emptying the cavern, till the feeding stream at o is sufficiently enlarged. So 

 that by these different constructions of the syphon, there may be some foun- 

 tains which shall fiow constantly in the winter, or a wet season, and intermit 

 in the summer, or a dry season ; and on the contrary, others which shall flow 

 continually in the summer, or a dry season, and intermit in the winter, or a 

 wet season. There is a third variety, which may arise from the make of the 

 syphon, and will occasion such irregularities as admit of no certain explanation. 

 This happens when the discharge of the syphon at the very last, is just equal 

 to the feeding-stream, and the civity of the syphon at n is large ; for in this 

 case, the air-bubbles, made by the fall of the feeding-stream from o to the 

 bottom of the cavern, will sometimes accidentally get into the mouth of the 

 syphon at m, and lodging at n, will so choke it, as to render its running and 

 stopping, as well as the quantity of its discharge, entirely uncertain ; so that 

 these sort of fountains will admit of no further consideration. 



But before leaving the consideration of fountains explicable by one reservoir 

 and syphon, it may not be amiss to observe, that those which intermit regularly 

 will have their flux always longer, and their pause or intermission shorter, in 

 winter and in wet weather, than in summer or in a dry season; which is a con- 

 sequence of this hypothesis, by which it may be examined, whether it be ap- 

 plicable to any particular intermitting fountain or not. 



If the single reservoir and syphon have another outlet at r, fig. 11, situated 

 between the bottom cd of the cavern, and the top of the syphon n, we shall 

 have another kind of fountains. For if the feeding- stream at o, be capable of 

 being discharged by the outlet at r, a fountain derived from r will continually 

 run, while the feeding-stream can be discharged that way, and will increase 

 and decrease with any little alteration happening to the feeding-stream at o, 

 provided the said stream does not grow too large for the outlet at r. But in 

 that case the cavern must be tilled up to n, and the syphon may begin to play; 

 which, together with the outlet at r, may discharge so much, as to make the 

 surface of the water in the cavern sink below k, and consequently the fountain 

 proceeding from r must stop. If the discharge of the syphon be so great as to 

 4 A 2 



