occurred in Lima in 1791, was 78°.4, and exceeded that of the hottest month of the same year in London, August, by 9°.4. The March of the following year was still hotter in Lima, the mean degree of temperature having been 80°.25. The cause of the difference of more than 9° of mean temperature between the two capitals, at the same time that the mercury was never higher at Lima than it has been in London, appears to have been owing to the constancy of the heats in the former city, and to the few variations to which the atmosphere is there subjected. Between the 17th and 30th of September 1791, both days included, there was not the slightest variation in the observations made at noon. In London, on the other hand, the fluctuations, which are at all times unceasing, produced in the month of June 1791, according to the observations of the afternoon, varieties of temperature to the extent of 28°. But for the presence of the southerly winds which prevail on the coast, the changes of temperature at Lima would be still less, as in other parts of Peru. In the plains of Bombon, there is a variation of 6° only, throughout the year; and at Santa Fe, the observations of two successive years afforded a result of only 30 of variation.
To complete the series of subjects introduced into this work, an Indian woman of a village near Lima, holding in her arms an infant, is represented in Plate XX. in her simple and rustic garb. The portraiture of this female peasant maybe regarded as a companion to the male Indian wearing the poncho, delineated in Plate XVII.
APPENDIX.
