The plates E. 3 & N. 1 were also varnished before then and dried before fire.

Sat. Nov. 7 [1772]. N. 5 and E. 8 were varnished with rosin, and N. 6 and E. 6 varnished with 8 parts of solut. rosin & 3 of vermilion, and then baked for about 2 hours with heat which part of the time rose to 146, but commonly did not exceed 130.

Sun. Nov. 7 [1773*]. Th. 63. Com. - 2.

N. 3 closed and sep. wide immed.

N. 8 closed & sep. again in 2" or 3′′.

N. 6 rather slower.

N. 5 closed and sep. wide immed.

N. 4 more so.

N. 2 D.

N. 1 D°.

E. 6 closed in 3 or 4".

N. 221

E. 8 closed and sep. again in 2 or 3" about 1 inch.
E. 3 closed and sep. wide immed.

4, 2, 1

rosin and verm. 1st

(rosin alone, 2 1st

Th. Nov.

4 soft cement

E

6 rosin and verm. last done

5 hard cem.

8 rosin last done

1 rosin alone

2 & 3 ros. and verm. 1st ros. alone 1*

4.21 of the lac varnish contains 12 gra. of lac.

N. 8 in soft cem. closed almost immed. sep. again in 3 or 4".
E. 4 in Do did not quite close in min.

E. 8 in ros. closed and sep. again almost immed.

E. 6 ros. and verm. closed and sep. wide immed.

N. 5 ros. more so.

N. 6 ros. & verm. same as E. 6.

Th. Nov.

Th. 58. Com. 7. N. 21.

E. 4 at first approached a little nearer, afterwards did not.

N. 8 closed and sep. again in about 3".

E. 2 Lac and verm. closed and sep. wide immed.

E. 3 Lac. not so soon.

N. 2 Lac and verm. rather quicker.

N. 1 Lac. Do.

* [Probably Nov. 8, 1772. See Note to Art. 502.]

E. 1 not varn. closed in about 8".

N. 4 closed and sep. again in about 2′′.

499] Sat. Nov.

E. 8 & N. 5 were varnished with ros. & E. 6 & N. 6 were varnished with 8 parts of solut. rosin & 3 of vermillion. They were afterwards baked over boiling water for about 2 hours, the heat between 115 & 120.

Sun. Nov.

E. 1 & N. 4 were varnished with 6.0 of thick solut. lac, 3.3 of verm. & 19 of spts, and E. 2 & N. 1 were varnished with 60 of solut. lac, 4·16 of verm. & 9 of spts. last mixture spread on the glasses was 8.0.

Mon. Nov. 16 [1772]. Th. 52.

The quantity of this

N. 2 not covered, closed rather sooner and sep. rather more than

Two trial plates were made with two plates of glass with rosin between, for comparing thick rosin with the double plates A and B.

Two trial plates were also made on a piece of the white plate glass for comparing N and thin white with D + E.

501] Sat. Oct. 17 [1772]. Th. 65. N. 21. C. 7.

Plates tried.

D of Nairne

varnished with lac

G of N in cement]

70

D°

А

D°

1.58

Hence it should seem that the plate D contained about 1.6 inc. el. less than the plates E or G, which is nearly conformable to p. 26, 1772 [Art 489].

The thick rosin plate seems to contain just the same as doub. pl. B, and the 2nd rosin plate to contain 2.49 inc. el. less than D. The thick white seemed to contain 79 inc. el. less than D.

502] Q and P compared with M and K of Nairne, also green cylinder L and white cylinder compared with plates of Nairne by means of sliding trial plates.

Mon. Oct. 18*. Th. 64. N. 174. C.-15.

[13 observations, Art. 660.]

[Result.] Therefore K seems to contain 5 inc. el. less than M, conformable to 1772, p. 26 [Art. 489].

Q contains 16 inc. less, and P 16 less.

The comp. power of white cyl. = 537·5, and [it] appears to contain 756 inc. el. Therefore inc. el. by comp. power = 1·41.

* [As the records of the actual observations in the following articles are of precisely the same nature as those already given, they will be omitted, and as the author has summed up the results for each day, these statements only will be given, except in cases of more than ordinary importance. According to the day of the week and month the dates for these experiments should belong to 1773, but as the experiments seem continuous with those of dates before and after which are certainly in 1772, I think Cavendish made a mistake in the day of the month which he did not find out till 4th November.]

The comp. power of green cyl. is 318-2, and [it] appears to contain 540 × 1 inc. el., therefore inc. el. by comp. power 1.62*.

22

503] 1st and 2nd green and white cylinder and white jar compared with H of Nairne in usual manner.

[12 observations.]

[Result.] Hence it should seem that the white cyl. contained as much el. as H; the 2nd green contained 45 inc. el. more than H; the 1st green uncertain, and the white jar seemed to contain 74 inc. el. more.

504] Trials of the same cylinders and jar in same manner except that in trying the white jar and 18 green cylinder the plate M of Nairne was placed on the neg. side as an additional trial plate.

Sat. Dec. 5 [1772]. Th. 56. N. 20.

[13 observations. Art. 660.]

[Result.] Hence 1st green cylinder should cont. 135 inc. el. more than H

By means of this and preceding page, the quant. el., comp. power and quant. el. by comp. power are as followst:

505] The quant. el. in the 2 coated globes was tried by putting the white cylinder and the 6th sliding plate on neg. side.

[6 observations.]

[Result.] Therefore globe 2

globe 3

seems to contain 1555 [circ.] inc. el.

(1782

Trials of jars used in the 1st sort of experiments: [Art. 240] tried by putting a sliding plate with or without the white cylinder on neg. side.

Th. Dec. 3 [1772]. Th. 55. N. 22.

[6 observations.]

[Result.] There seems some mistake in the 3rd exper., therefore if we make use only of those exper. in which they sep. pos. the jar for (neg. side

pos. side

162
0

contains more than H;

* [These measures of specific inductive capacity must be multiplied by 5·1. See

note to Art. 462.]

+ [See Art. 383.]

inc. of el., but if we made use only of the other exper. it

would be neg. side 1233

side.

pos. side 1043*

506] Trials of the 4 large jars, the jars being placed on the neg.

507] Trials of the 5th and 6th trial plates, the trial plate being placed on neg. side.

Result.

Therefore trial plate 6-56=H. [Trial plate]

6-18 rather more than C.

By mean 6-18= C, and trial plate 5 – 17 = C.

The trial plate 4 is on same plate as 5, and the area of one division on it is to that of 5 on trial plate 5 as 1·82 to 9.

508] Thick white, 2nd rosin, D and F of Nairne, and the two double plates together, compared together, also thin white with D and E and D and F.

Sat. Dec. 5 [1772] in evening. Th. 56. N. 284. [24 observations.]

N.B. Before these exper. were tried, the plates E and F were freed from cem[ent] and coated afresh with plates of same size. The plate D was also freed from the varnish and coated afresh, and the trial plate B was freed from cement and coated with rather larger plates.

Hence it should seem that thick ros. contained 11 inc. el. less than the doub. plates A or B, id est 18.2 inc. el., that the thick white contained same as D, id est 36 inc., and that 2nd ros. contained 2.03 inc. less, id est 34 inc., which differs very little from p. 12 [Art. 501].

M.

509] Whitish plate, P, Q, O, old G and thin rosin compared with