VOL. LIII.j PHILOSOPHICAL TRANSACTIONS. & 



manner. 1st. He ordered one of his negroes to gather him a pint of those 

 berries, from which he extracted almost |-of a pint of a juice, and boiled it with 

 a pint of Bristol water, -j^ of an hour. 2dly. He then took 2 pieces of flannel, 

 and numbered them 1 and 2, boiled them in a separate tin-pot with alum J- of 

 an hour, and rinced them in cold water. 3dly. He then dipped the piece of 

 flannel N° 1 into the pot where the juice was, and left it to simmer 5 minutes; 

 then took it out, and rinced it in cold water, when, to his surprise, he found a 

 superior crimson dye fixed on the flannel than the juice of the berry. 4thly. 

 He then dipped the piece of flannel N" 2 in the same juice, and being desirous 

 to clean his hands from the stain which N" 1 had caused, he ordered some lime 

 water to be- brought him, such as he used to settle the indigo, and found the co- 

 lour of the stain change to a bright yellow. This unexpected change urged him 

 to throw a wine glass full of lime water into the pot, where the piece of flannel 

 N" 2 was simmering; on which all the juice, as well as the flannel, became of 

 a bright yellow; by which he found alum fix the crimson, and lime the yellow. 

 5thly. Having then put a quart of fresh juice in 2 pint decanters, in one of 

 which he put a small quantity of powdered alum, he laid them up; about 6 

 weeks after he examined them, and found the juice in the decanter, which had 

 no alum, was turned black, and the other retained its colour. 



XXXFIII. On the Eclipse of the Sun, Jpril 1, 1764. % Mr. James Fer- 

 guson, F. R. S. p. 240. 



This was a projection of the eclipse of the sun, which was to happen on the 

 1st of April 1764. The diagram shows the time and phases of that eclipse, for 

 the Royal Observatory at Greenwich, and the calculation is from Meyer's tables. 

 As these tables gave the appearance and the times very different from those of 

 Flamsteed, Halley, and La Caille, Mr. F. oflfers this projection as a means of 

 proving which tables are the more accurate. 



If the motions of the sun and moon were equable, any given eclipse would 

 always return in a course of 223 lunations, which would consist of 18 years 11 

 flays 7 hours 43 minutes 20 seconds (as was observed by the ancients) for 1388 

 years; and would for ever do so, if, at the end of each period, the sun and moon 

 should be in conjunction either in the same node, or at the same distance from it 

 as before. But that is not the case: for if the sun and moon are once in conjunc- 

 tion at 18" distance from the node, which is the greatest distance at which the 

 moon's shadow can touch the earth, at the next period of 18 years J 1 days, &c. 

 the sun and moon will be 28' 12" of a degree nearer the same node than they 

 were at the period last before. And so by falling gradually nearer and nearer the 

 same node every time, the moon's shadow will pass over the centre of the earth's 

 enlightened disk, at the end of the 38th periodical return of the eclipse, from the 

 time of its first coming in at either of the earth's poles; because the conjunction 



