658 ULLOA's voyage to south AMERICA. 



continued in the harbour till the tempeft was over ; and he affured me that his conform- 

 ing to this obfervation had been of the greateft advantage to him ; relating, in confirm- 

 ation of his opinion, that being once at Baldivia, the governor, fo far from regarding his 

 apprehenfions from fuch prefages, turned them into ridicule, and infilled on his putting 

 to fea, but was foon convinced that thefe omens were not chimeras ; for the veffel was 

 hardly out of the harbour, when a ftorm at north came on with fuch violence, that it 

 was with the utmofl difficulty fhe was faved from being wrecked in that bay ; and this 

 would infallibly at laft have been the confequence, had the ftorm continued fome time 

 longer ; for, even when the wind abated, they found it hardly poflible to carry her into 

 the harbour to repair the damages Ihe had received. 



Other obfervations relating to thefe northerly winds are, that they always blow when 

 the Sures are in their ftrength, in the higher latitudes, and alfo between the parallel of 

 20^ and that of Panama, it being then winter in thofe climates j and are alfo found in 

 latitudes beyond 20^, but never nearer to the equinoctial. Another obfervation is, that 

 during the time of the brifas, between Panama and the equinoftial, thefe winds are never 

 felt in any part of the Pacific Sea, the fouth winds alone prevailing there. Laftly, it is 

 obferved, that within thirty or forty leagues of the coaft of Chili, while one part is agi- 

 tated with ftorms at north, the fouth winds frefhen in another. This, however fingular 

 it may appear, is no more than what was experienced by the three fhips, Efperanza, 

 Belen, and Rofa, which being at the mouth of the Bay of Conception, the latter took 

 her leave of them, and bore away with a frefh gale at fouth to Valparaifo, whilft the 

 others who fteered for the iflands of Juan Femandes, were overtaken in their paflage by 

 a ftorm at north. 



As in fummer the fouth winds generally fhift between the fouth-fouth-eaft and eaft- 

 fouth-eaft, fo in winter they continue for fome time between the fouth-weft and fouth ; 

 confequently there is a neceffity, in the latter feafon, to ftand out to fuch a great diftance 

 from the coaft in queft of them, as muft be done in fummer. 



CHAP. IV. — Account of ihe IJlands of Juan Femandes, Voyage from thofe Iflands to 

 Santa Maria, and from thence to the Bay of Conception, 



THE iflands of Juan Femandes, which, on account of their fituation, belong to the 

 kingdom of Chili, are two in number ; one, as lying farther to the weft, is diftinguifhed 

 by the epithet De Afuera ; and the other, as nearer the land, or to the eaftward, is 

 called La de Tierra. The former, which is fomething above a league in length, is 

 nearly of an oval figure, and the land very high, fo that it has the appearance of a round 

 mountain ; and its fteepnefs on all fides renders it every where almoft inacceffible. Se- 

 veral large cafcades tumble from its fummit, and the water of one of them, after a fuc- 

 ceffion of long falls among the rocks on the fouth-weft fide of the ifland, precipitates 

 itfelf into the fea with fuch amazing impetuofity, that its froth may be feen at three 

 leagues diftance. The longitude of this ifland, according to the reckoning of Don 

 George Juan, admitting the currents to fet towards the fouth-weft, is 3° 20' W. from 

 the meridian of Callao j but, according to my computation, 3^ 27'. By the coaft we 

 fteered from the meridian De Afuera till we reached La de Tierra, we concluded the 

 diftance between thofe iflands to be thirty-four leagues. 



The ifland De Tierra, which is four hundred and forty leagues to the north of Cape 

 Horn, is between three or four leagues from eaft to weft, which is its greateft length. 



9 It 



