BOOK II. OF THE ATMOSPHERE. 265 



1276. The situations of the moon favorable to bad weather are the perigee, new and full moon, passage of 

 the equator, and the northern lunistice. Those belonging to the former are, the apogee, quadratures, 

 and the southern lunistice. Changes of the weather seldom take place on the very days of the moon's 

 situations, but either precede or follow them. It has been found by observation, that the changes- 

 affected by the lunar situations in the six winter months precede, and in the six summer months follow 

 them. 



1277. The octants. Besides the lunar situations to which the above observations refer, attention must be 

 paid also to the fourth day before new and full moon, which days are called the octants. At these times the 

 weather is inclined to changes ; and it may be easily seen, that these will follow at the next lunar 

 situation. Virgil calls this fourth day a very sure prophet. If on that day the horns of the moon are 

 clear and well defined, good weather may be expected ; but if they are dull, and not clearly marked on the 

 edges, it is a sign that bad weather will ensue. When the weather remains unchanged on the fourth, 

 fifth, and sixth day of the moon, we may conjecture that it will continue so till full moon, even sometimes 

 till the next new moon ; and in that case, the lunar situations have only a very weak effect. Many 

 observers of nature have also remarked, that the approach of the lunar situations is somewhat critical for 

 the sick. According to Dr. Herschel, the nearer the time* of the moon's entrance, at full, change, 

 or quarters, is to midnight (that is within two hours before and after midnight), the more fair the weather 

 is in summer, but the nearer to noon the less fair. Also, the moon's entrance, at full, change, or 

 quarters, during six of the afternoon hours, viz. from four to ten, may be followed by fair weather ; but 

 this is mostly dependent on the wind. The same entrance during all the hours after midnight, except the 

 two first, is unfavorable to fair weather ; the like, nearly, may be observed in winter. 



1278. The artificial data are the barometer, hygrometer, rain-gauge, and ther- 

 mometer. 



1279. By means of the barometer, Taylor observes, we are enabled to regain, in some 

 degree at least, that foreknowledge of the weather, which the anciertts unquestionably 

 did possess ; though we know not the data on which they founded their conclusions. 

 We shall therefore annex such rules, as have hitherto been found most useful in ascer- 

 taining the changes of the weather, by means of the barometer. 



1280. The rising of the mercury presages, in general, fair weather; and its falling 

 foul weather, as rain, snow, high winds, and storms. 



The sudden falling of the mercury foretels thunder, in very hot weather, especially if the wind is 

 south. 



The rising in winter indicates frost ; and in frosty weather, if the mercury falls three or four divisions, 

 there will follow a thaw : but if it rises in a continued frost, snow may be expected. 



When foul weather happens soon after the falling of the mercury, it will not be of long duration ; nor are 

 we to expect a continuance of fair weather, when it soon succeeds the rising of the quicksilver. 



If, in foul weather, the mercury rises considerably, and continues rising for two or three days before the 

 foul weather is over, a continuance of fair weather may be expected to follow. 



In fair weather, when the mercury falls much and low, and continues falling for two or three days before 

 rain comes, much wet must be expected, and probably high winds. 



The unsettled motion of the mercury indicates changeable weather. 



1281. Respecting the words engraved on the register-plate of the barometer, it may be 

 observed, that they cannot be strictly relied upon to correspond exactly with the state of 

 the weather ; though it will in general agree with them as to the mercury rising and 

 falling. The words deserve to be particularly noticed when the mercury removes from 

 ' changeable' upwards ; as those on the lower part should be adverted to, when the mer- 

 cury falls from ' changeable' downwards. In other cases, they are of no use : for, as its 

 rising in any part forebodes a tendency to fair, and its falling to foul weather, it follows 

 that, though it descend in the tube from settled to fair, it may nevertheless be attended 

 with a little rain ; and when it rises from the words < much rain' to rain' it shows only 

 an inclination to become fair, though the wet weather may still continue in a less consi- 

 derable degree than it was when the mercury began to rise. But if the mercury, after 

 having fallen to 'much rain,' should ascend to 'changeable,' it foretels fair weather, 

 though of a shorter continuance than if the mercury had risen still higher ; and so, on 

 the contrary, if the mercury stood at ' fair' and descends to < changeable,' it announces 

 foul weather, though not of so long continuance, as if it had fallen lower. 



1282. Concavity of the surface of the mercury. Persons who have occasion to travel 

 much in the winter, and who are doubtful whether it will rain or not, may easily ascer- 

 tain this point by the following observation : A few hours before he departs, let the 

 traveller notice the mercury in the upper part of the tube of the barometer; if 

 rain is about to fall, it will be indented, or concave ; if otherwise, convex or pro- 

 tuberant. 



1 283. Barometer in spring. Towards the end of March, or more generally in the be- 

 ginning of April, the barometer sinks very low, with bad weather ; after which, it seldom 

 falls lower than 29 degrees 5 minutes till the latter end of September or October, when 

 the quicksilver falls again low, with stormy winds, for then the winter constitution of the 

 air takes place. From October to April, the great falls of the barometer are from 29 

 degrees 5 minutes to 28 degrees 5 minutes, and sometimes lower ; whereas during the 

 summer constitution of the air, the quicksilver seldom falls lower than 29 degrees 5 

 minutes. It therefore follows that a fall of one tenth of an inch, during the summer, 

 is as sure an indication of rain, as a fall of between two and three tenths is in the 

 winter. 



1284. Barometer relative to situation. It must, however, be observed, that these 

 heights of the barometer hold only in places nearly on a level with the sea; for cxpe~ 



