CONTRACTION OF WATER BY HEAT. 363 



ration oP its furface by a cylindrical folid, is liable to the ob- change in the 

 jection that it fuppofes the attraction or repulfion between the puffion'bctweeii 

 folid and the fluid to remain unchanged by variations of tern- the water and 

 perature; whereas the contrary feems moft probable. The* ^^™^ thc 

 doctor's experiment mud be grounded upon a pofition that the 

 greater the depreffion or the greater the elevation of a fluid 

 round a fmall cylinder partly immerfed in it, the greater muft 

 be the fefiftance from imperfect fluidity. But thefe effects are 

 evidently as much governed by'the attraction or repulfion of 

 the folid with regard to the fluid as by the refinance which the 

 experiments are intended to meafure. I have fomewhere read Water clock- 

 that water clocks and other inftruments for meafuring time, flower in cola 

 by the paffage of water through fmall holes, go flower in cold weather; be- 

 wealher. This may arife from contraction of the hole, though i^tfsfru'id* * 

 my author afcribes it to imperfect fluidity. After fome medi- 

 tation on this problem it (till appears to me to be furrounded 

 with difficulties. Perhaps it may be one of the beft methods 

 to fuffer the fluid to drop from a capillary fyphon in different 

 temperatures. I am difpofed to think that the drops would be Suppofition that" 

 fmalleft and the whole quantity in a given time greateft wn en™ ft f lterwou!4 



1 J ° ° drop from a ca- 



the fluidity was the mod perfect, or at leaf! when the adhefion pillary tube wheo 

 of the particles of the fluid to each other was the Ieafh But " ^ as moft 

 even here the attraction of the fmall capillary extremity of the , 

 tube from which the drop would fall would require to be con-' 

 fidered; and on this account the method would be preferable 

 (if fo) to Dr. Hope's only becaufe the repetition of u great 

 number of drops or quantity of effluent water would give a 

 greater degree of precifion to the refult. 



Is it likely that the rope pump turned regularly a certain Will the rope 

 number of turns in a given time would raife more water when P U ™P fte ™ a t 



. , difference in the 



coldeft and leaft fluid? If it did not might we not infer that tenacity or flu- 



the fluidity of water is not fenfibly affected by change of tern- ! dity of ***? 



J J . r hot or cold } 



perature ? 



Obfirvations 



