304 PHILOSOPHICAL TRANSACTIONS. [aNNO 1686. 



an sooth part of its space, it would be equally dense with water, which yields 

 not to any force whatever, as has been found by several experiments. Nor can 

 the rarefaction proceed in infinitum ; for, supposing the spring, by which it di- 

 lates itself, to proceed from what texture of parts you please, yet there must be 

 a determinate magnitude of the natural state of each particle, which they cannot 

 exceed, when freed from all manner of pressure. 



Though these objections affect the strict accuracy of these conclusions, drawn 

 from the specific gravity of the air, observed at any time ; yet the method, here 

 shown, will by a like calculation give the heights of the quicksilver, and the 

 refractions of the air, from any assigned height of the barometer at the earth's 

 surface, and any specific gravity given. As to the condensation and rarefaction 

 by heat and cold, and the various mixture of aqueous and other vapours, these 

 two objections seem generally to compensate each other ; for when the air is 

 rarefied by heat, the vapours are raised most copiously ; so that though the air, 

 properly so called, be expanded, and consequently lighter, yet its interstices 

 being crouded full of vapours of much heavier matters, bulk for bulk, the 

 weight of the composition may continue much the same. 



Having mentioned the difference there is between the height of the mercury 

 at one time, from that at another, it may not be unacceptable to offer some 

 reasons for the said difference, which seem to have some appearance of truth. 

 First then, it is demonstrable that the height of the cylinder of mercury, is 

 equal to the weight of the whole incumbent air ; and consequently that differ- 

 ence in the height of the mercury must be owing either to a decrease or in- 

 crease of new matter to the air. That hypothesis therefore that shows how the 

 air shall be increased or diminished in any particular place, will give a reason 

 for the greater and lesser height of the mercury in the baroscope. But to direct 

 us in the choice of the several causes, which may be assigned for the increase 

 and decrease of the air, it will not be unnecessary to enumerate some of the 

 principal observations made on the barometer. 



The first is, that in calm weather, when the air is inclined to rain, the mer- 

 cury is commonly low. 2. That in serene good settled weather, the mercury 

 is generally high. 3. That on very great winds, though they be not accom- 

 panied with rain, the mercury sinks lowest of all, with respect to the point of 

 the compass the wind blows upon. 4. That, casteris paribus, the greatest 

 heights of the mercury are found on easterly and north-easterly winds. 5. That 

 in calm frosty weather the mercury generally stands high. 6. That after very 

 great storms of wind, when the quicksilver has been low, it generally rises again 

 very fast. 7- That the more northerly places have greater alterations of the 

 baroscope, than the more southerly. 8. That within the tropics, and near 

 them, those accounts we have had from others, and my own observation at St. 



