At Oneeheow they brought us several large roots of a brown colour, shaped like a yam, and from six to ten pounds in weight.  The juice, which it yields in great abundance, is very sweet, and of a pleasant taste, and was found to be an excellent substitute for sugar.  The natives are very fond of it, and use it as an article of their common diet; and our people also found it very palatable and wholesome.  We could not learn to what species of plant it belonged, having never been able to procure the leaves; but it was supposed, by our botanists, to be the root of some kind of fern.

Agreeably to the practice of Captain Cook, I shall subjoin an abstract of the astronomical observations which were made at the observatory in Karakakooa Bay, for determining its latitude and longitude, and for finding the rate and error of the time-keeper.  To these are subjoined the mean variation of the compass, the dip of the magnetic needle, and a table of the latitude and longitude of the Sandwich Islands.

The latitude of the observatory, deduced
  from meridian zenith distances of the
  sun, eleven stars to the south, and four
  stars to the north of the zenith 19 deg. 28’ 0” N.
The longitude of the observatory, deduced
  from 253 sets of lunar observations;
  each set consisting of six observed
  distances of the moon from the
  sun or stars; 14 of the above sets were
  only taken at the observatory, 105 sets
  being taken whilst cruising off Owhyhee,
  and 134 sets when at Atooi and
  Oneeheow, all these being reduced to
  the observatory, by means of the timekeeper 204 deg. 0’ 0” E.
The longitude of the observatory, by the
  time-keeper, on the 19th January,
  1779, according to its rate, as found
  at Greenwich 214 deg. 7’ 15’ E.
The longitude of the observatory, by the
  time-keeper, on the 19th January,
  1779, according to its rate, corrected
  at different places, and last at Samganoodha
  Harbour, in Oonalaschka 203 deg. 37’ 22” E.
The daily rate of the time-keeper losing
  on mean time, was 9",6; and, on the
  2d February, 1779, it was 14^h 41’ 1”
  too slow for mean time. 
The variation of the compass, by azimuths,
  observed on shore with four
  different compasses 8 6 0 E.
The variation of the compass, by azimuths,
  observed on board the Resolution,
  with four different compasses 7 32 0 E.
Dip of the north Balanced needle\ 40 22 30 E.
  pole of the magnetic | |
  needle on | Unbalanced, or |
  shore, with \ plain needle 40 41 15 E.
Dip of the north Balanced needle\ 41 50 0 E.
  pole of the magnetic | |
  needle on | Unbalanced | 40 30 5 E.
  board, with \ needle