PACIFIC AND BEERING’S STRAIT. 
439 
the channel between Assumption and Agrigan about forty miles wide : CHAP, 
the Jesuits extend it to sixty ; but this cannot be the case, as it would 
place Affriffan near the latitude of 18° 45' N. in which parallel \bar- April, 
gottia, according to Espinosa, has placed the island of Pagon. It seems 
necessary therefore to contract the channel between Assumption and 
Agrigan as marked in the Jesuits’ plan, and to reduce the size of Agrigan 
in order to reconcile the position of the islands. Arrowsmith has in- 
correctly placed the Mangs on the south side of Assumption ; by our 
astronomical bearings they are situated IST. 27° 7' 30" W. (true) from the 
south-east end of that island, and are in latitude 19° 57' 02" N. They 
consist of three high rocks, lying in a south-easterly direction*. 
From what I saw of the island of Assumption it appears to be a 
very proper headland for ships coming from the eastward and bound to 
Canton to steer for. It is high, and may be safely approached in the 
night if the weather is clear ; and there is a wide channel to the south- 
ward of it. It is far preferable to adopt this channel than to pass to the 
northward of the Mariana group, which is sometimes done ; as I am cre- 
dibly informed that there is much broken ground in that direction. 
We have as yet no good chart of this group of islands. The geogra- 
phical position of Assumption and of the Mangs will be found in the 
table at the end of this work. 
Under the lee of the island we observed a great many birds, prin- 
cipally of the pelican tribe, of which there was a species supposed by 
our naturalists to be new. It is described as being smaller than the 
frigate-bird, and of a dark brown colour, with the exception of the belly 
and breast, w Inch were white, and the bill, which was either white or of 
a light lead colour. 
* It is somewhat remarkable that in passing to the southward of the island of Assumption, 
at the distance of four miles and a half, we did not discover the rocks which Captain Freycinet 
has supposed to be the Mangs, situated in latitude 19“ 32' N. Our latitude when in the 
meridian of Assumption was 19° 36' N. by which it is evident that we must have passed within 
four miles of these rocks, provided both latitudes he correct. Had I known of their existence 
at the time, I should certainly have stood to the southward, in order to connect them by 
triangulation with the Assumption and the Mangs; but Captain Freycinet’s discoveries were 
not then published. 
