taken by every perfon, who fhall obferve the fame phenomenon at one and the fame inftant of time; and thence the distance of Jupiter from the meridian of each will be known to feconds, if we fuppofe the latitudes known before-hand.

And if the telescopes of quadrants could be made fufficiently powerful to obferve the fatellites, then a fingle obferver, at any place, could perform the whole without trouble or difficulty, and would only need a common watch, and a little more patience than would be requifite if the watch were perfect and calculation

true.

But fuppofing the telescopes and quadrants as they are, and two obfervers at each place, one employed with the fatellite, and the other with the quadrant, then the latter must carefully keep the body of Jupiter on the line of altitude till the other tells him to ftop, which is to be done at the inftant of obferving the expected phe

nomenon.

By this mode a degree of longitude may be measured with as much accuracy as a degree of latitude; and it is what I have in contemplation to perform, as soon as I can get the requifite inftruments.

Remarks on some erroneous Observations of Jupiter's first Satellite.

In 1778 I took notice that when Jupiter is very near the oppofition, the observations are not to be depended on, and that the Satellite vanished without changing colour. The fame happened in 1779, 1780, and again in 1784, at Beemulwilsa; therefore I have put down the times of fuch obfervations, as they are reduced to apparent time, from the known deviation of the clock from mean time. The tranfit inftrument was examined VOL. I by

F

by stars that pass over the zenith, and by others north and fouth, and by equal altitudes; all which fhewed it was as nicely in the meridian as it well could be.

In 1779, on the 3d of March, I observed, and μ Gefe minorum, and the deviation was the fame as that derived from the tranfits of the fun on the 23d February and 5th March. In the observation of the moon 23d November following, the accuracy of its pofition was afcertained, fo that the times were correct; and the errors depend on fomething at the Satellite and planet. Perhaps Jupiter's atmosphere may be fo denfe as to prevent the free paffage of the diminished light foon after the beginning of an eclipfe, or even before it. If fo, thefe obfervations may tend to clear up that point, and to measure the extent of that atmosphere.

All these observations were made with Dolland's triple object glass.

Observations of Venus.

1776, 2d January, at 7h.55', in the morning, I meafured the diftance between Venus and the Sun 46° 32"

I was informed the natives were viewing it with aftonishment, but I did not fee it with the naked eye. Through the little telefcope of my HADLEY's quadrant it appeared as bright as Capilla.

1777, 1ft July, Venus paffed the meridian

14th, Venus visible to the naked

has been fo three days.

Paffed the meridian

Apparent Time corre&.

h,

21.30.41,5

eye, and

21.01.02,0

1780, 18th March, an appulse of Venus to Mars.

N. B. The fcale of the micrometer is divided into twentieth parts of an inch, and the nonius fubdivides these into twenty-five parts each.

1

Sun's diameter by Ephemeris 32′ 11′′,6, from which the distances were calculated.

Observations of the Moon.

1775, 12th January, an occultation of Aldebaran.

Immerfion

h

8.54.55

I believe

I believe the watch was fet by equal altitudes, but I have loft the book in which the entry was made, and have only a copy of my observations as a register of this and the next that follows.

15th February, an eclipse of the Moon.

End 10h.15'.00",5, apparent time correct.

1776, 3d March, an occultation of Regulus.

Not having an ephemeris at the time, the obfervation was accidental, and confequently not prepared for. The tranfit inftrument was but lately put up, and had not been much used, but it was the only resource for time accordingly, it was adjufted truly as to level and wires, but it was not in the meridian accurately. Therefore the tranfits of feveral ftars were taken to determine the position of the instrument, and the error of that being known, the times could be corrected by a very easy rule, which I fubjoin. Let x be the error in feconds at the horizon, a and b the fines of the zenith diftances of two stars, a and в the fines of the polar distances, d the difference of the errors of the clock, as found from the obferved and the calculated tranfit of those two ftars.

x a

A

Then will be the space at the equator for the equation to correct one, and the fame for the other;

B

b

and the fum of these two will be equal to d+15. or,

x b 15 A 15 B 15 + d + A B

a B + b A

d. Whence for

and, Leonis were the two ftars that were relied on for time and pofition, because they pass so nearly at equal distances from the zenith, that the mean of their errors of the clock would be fo near to the true one,

that