VOL. XXXVII.J PHILOSOPHICAL TRANSACTIONS. 54^ 



the whole time of the first flux and intermission, in comparison of the whole 

 time of the second flux and intermission, it is a particular, requiring so many 

 things to be taken into consideration, for determining it in each case, that he 

 contents himself with showing that it may be longer, by an experiment that 

 will presently be made. Another variety in this sort of fountains might be made 

 by a second feeding-stream z, coming into the second reservoir efgh; but the 

 bare mentioning of that will at present be sufficient. 



If in the contrivance of a single reservoir and syphon, the stream derived 

 from the syphon should fall into another reservoir ikkl, fig. 13, having no 

 syphon, but only a common outlet x, and should in this reservoir meet and 

 join with another stream constantly running, a fountain derived from the said 

 outlet X would be a reciprocating spring; by which name he calls those springs 

 which flow constantly, but with a stream subject to increase and decrease, to 

 distinguish them from intermitting springs, which flow and stop alternately. 

 And if the outlet x be too small to carry off" all the water brought into the 

 reservoir ikkl, by the syphon, over and above what is brought in by the con- 

 stantly running stream w; then the surface of the water in tlie said reservoir 

 ikkl must continually rise, till the velocity of the stream, going out at x, is 

 sufiiciently increased, to carry off" the water coming in; on which, the dis- 

 charge of the syphon being continually lessened, the said surface will again 

 subside, and the velocity of the stream at x will diminish; so that both the 

 increase and decrease in this reciprocating fountain will be gradual. Besides, if 

 the reservoir ikkl, or the channel derived from it, should have any leaks, cre- 

 vices, or other outlets, the water will issue through them on the rising of the 

 surface in the said reservoir, and occasion springs, which will cease again when 

 the surface subsides. 



Let us now suppose such a reservoir ikkl, fig. 14, with a constantly running 

 stream w, and an outlet x, to receive the water of a syphon stv, coming 

 through two reservoirs abcd and efgh, as before described. A fountain derived 

 from x in this case, would be an intermitting reciprocating spring, whose stream 

 would reciprocate; but its reciprocations would sometimes stop, and have fits 

 of intermission. 



Such, in all probability, is the fountain called Laywell, before described, 

 whose phaenomena gave occasion to these thoughts, and seem capable of being 

 accounted for by such a contrivance. And for the better discovery of the nature 

 of this fountain, whether it be owing to such a piece of natural machinery, or 

 otherwise, it would be proper to observe the length of time of each increase, 

 decrease, and pause, in every reciprocation, with the number of reciprocations 

 in every reciprocating fit, and also the length of the intermissions of the said 



