716 



SCIENCE. 



[N. S. Vol. IX. No. 259. 



astronomical and meteorological work of the 

 Observatory of Harvard College in Peru is the 

 result of a bequest left to the Observatory in the 

 will of Mr. "Uriah A. Boyden, in 1887. Under 

 the terms of the will this monej' was to aid in 

 the establishment of an observatory ' ' at such 

 an elevation as to be free, so far as practicable, 

 from the impediments to accurate observation 

 which occur in observatories now esisting, 

 owing to atmospheric influences." It was in 

 connection with the study of the atmospheric 

 conditions of the desert strip of the west coast of 

 South America, with a view to determining the 

 best possible site for the new observatory, that 

 the early meteorological observations in Peru 

 were undertaken. The stations selected for the 

 taking of these preliminary observations were 

 Mollendo, Arequipa, Vincocaya, Puno and 

 Chosica. The first four stations are between 

 latitude S. 15° 40' and S. 17° 5', on the 

 Ferrocarril del Sur del Peru, which runs from 

 Mollendo, on the sea coast, northeast to Puno, 

 on Lake Titicaca, a distance of 325 miles (by 

 rail). Mollendo is immediately on the coast (alti- 

 tude 80 feet). Arequipa is at a distance of 80 

 miles in a direct line from the Pacific Ocean, at an 

 altitude of 7,550 feet. Vincocaya is 14,360 feet 

 above sea level, on a desolate plateau, near the 

 crest of the Western Cordillera. Puno (12,540 

 feet) is on the western shore of Lake Titicaca. 

 The station at Chosica was situated about 25 

 miles northeast of Lima (altitude 6,600 feet). 

 A few observations, chiefly of cloudiness, were 

 made at the Pampa Central, near the central 

 western part of the Desert of Atacama, in 

 Chile. 



These early observations were made during 

 the years 1888-1890, with more or less com- 

 pleteness. They are, however, preliminary. 

 They were almost all made by observers who 

 had had little or no experience and who re- 

 ceived no compensation for their services. The 

 instrumental equipment in use at the different 

 stations vaa-ied considerably ; the hours of ob- 

 servation were not alwaj's the same ; the loca- 

 tion of the instruments was sometimes changed. 

 In short, the work as a whole was done in an 

 unsystematic and incomplete and often in a very 

 inaccurate way. This was, of course, abso- 

 lutely unavoidable. It was impossible to secui'e 



trained observers, to inspect the stations, or to 

 test the instruments. The observations were, 

 therefore, liable to be considerably in error. 

 Thus, in connection with the minimum ther- 

 mometer readings at Chosica the statement is 

 made in a note that it is probable that the 

 lower end of the index in the minimum ther- 

 mometer was read, instead of the upper end. 

 And in the wind observations at Arequipa and 

 Vincocaya it is noted that "the direction of the 

 wind was always given, even if the remark 

 appended was 'calm' or 'dead calm.' Ap- 

 parently the position of the wind-vane was re- 

 corded, whether at the time wind was observed 

 or not." These two cases will serve to indicate 

 the sort of errors which inevitably appear in 

 these records. We do not intend to criticise 

 adversely the publication of these early Peru- 

 vian observations, but merely to point out their 

 necessary inaccuracies. Professor Pickering says 

 very clearly in the preface: ' ' These observations 

 must not be regarded as indicating the accuracy 

 of those made later. * * * It must be remem- 

 bered that it was not possible under the condi- 

 tions then existing to obtain observations of 

 the accuracy of those made by professional ob- 

 servers at permanent and easily accessible ob- 

 servatories." And again, on page 68, Professor 

 Bailey says: " The results are perhaps as reliable 

 as are possible in such outlying stations, where 

 experienced observers cannot be obtained and 

 frequent supervision is impossible." 



The published observations comprise twenty- 

 nine tables. The data are by no means equally 

 complete for all stations. At Mollendo, Are- 

 quipa and Vincocaya the instruments in use- 

 were the maximum, minimum and ordinary 

 thermometers, thermograph and rain-gauge. 

 At Arequipa a solar radiation and a wet-bulb 

 thermometer were also used. At Puno the ob- 

 servations were continued but a short time, 

 and there was no thermograph. At the Cho- 

 sica station, in addition to the above-named in- 

 struments, there were a barograph, sunshine and 

 pole-star recorder. At Pampa Central cloud 

 observations only were made, four times daily. 

 There are several tables showing the hourly 

 means of the barograph and thermograph, and 

 a comparison of thermometer and thermograph 

 hourly and monthly means. Curves are also 



