488 don’t trust too much to their accuracy. 
up steam and anchor and continued on our way for the 
mouth of the Anioorj through a storm of rain, wind, and 
drifting fog-hanks that soon rendered our navigation so 
uncertain and dangerous that we were glad to gain an 
anchorage under the sheltering heights of another of the 
group until the return of clear weather. The next day 
it had improved considerably, and we again got under 
steam and ran along down the coast, hoping to be able 
to pass through between the island of Sagulien and the 
mainland into the Gulf of Tartary, and, following the 
w’est coast of Sagalien, finally get out into the Pacific 
through the Straits of La Perouse and then shape our 
course for San Francisco. In tliis, however, we were 
destined to fail, as the reader wdll see. 
AVhile at Ayan we had been presented by “old Fry- 
bark” with a large number of Russian charts, several of 
which related to the mouth of this river. These several 
charts embodied the latest surveys of that region, but 
were given to us by Mr. Freighburg with the warning 
remark, “Don’t trust too much to their accuracy, for the 
channels there sounded out have long since been filled in 
by shifting sand-banks, and even 'WC have to trust entirely 
to pilots.” We therefore proceeded with our usual cau- 
tion, and, while thus feeling our way with lead and look- 
out over the unknown ground, we were forced to anchor 
at night and continue with I'eturning light the previous 
day’s work. It was a running survey that we were 
making of this coast, and we had consequently to steam 
well in along the laud in order that Hartman might 
sketch in the coast-line as we proceeded. We found the 
water shoaling perceptibly as we neared the latitude of 
