23-^ COLOURED CLASSES OF THE ANCIENTS. 



might have the appearance of a groundlefs afTertion." After 

 which he mentions a floor formed of green glafs-plates difco- 

 vered in the Farnefe-ifland, as well as fome fragments of 

 glafs-cups, which muft have been turned on the lathe, and 

 then proceeds as follows : *' But the art ftrongly claims our 

 admiration in two fmall pieces of glafs, which laft year (1765^ 

 were brought to Rome. Each of them is not quite one inch 

 Very curious long, and one-third of an inch broad. One plate exhibits, on 

 of a duck} ^ ^^^^ ground of variegated colours, a bird reprefenting a 

 duck of various very lively colours, more fuitable to the Chi- 

 nefe arbitrary tafte, than adapted to (liew the true lints of na- 

 ture. The outlines are well decided and (harp, the colours 

 beautiful and pure, and have a very ftriking and brilliant 

 effedl; becaufe the artift, according to the nature of the parts, 

 has in fome employed an opake, and in others a tranfparent 

 traced and fi- glafs. The moft delicate pencil of the miniature painter 

 treme accuracy could not have traced more accurately and difiindly, either 

 aod effeft ; the circle of the pupil of the eye, or the apparently fcaly fea- 

 thers on the breaft and wings, behind the beginning of which 

 and continued this piece had been broken. But the admiration of the be- 

 Kvhokthkkntfsof^'^*-^^^'^^ 's at the higheft pitch, when, by turning the glafs, he 

 the piece, mi- fees (he fame bird on the reverfe, without perceiving any dif- 

 Tttb'^furfaw' ""ference in^the fmalleft points ; whence we could not but con- 

 clude, that this pi6tui»e is continued through the whole thick- 

 nefs of the fpeclmen ; and that, if the glafs were cut tranf- 

 verfely, the fame pifture of the duck would be found repeated 

 in the feveral flabs ; a conclufion which was ftill farther con- 

 firmed by the tranfparent places of fome beautiful colours upon 

 the eye and breaft that were obferved. The painting has on 

 both fides a granular appearance, and feems to have been 

 formed, in the manner of mufaic works, of fingle pieces; but 

 fo accurately united, that a powerful magnifying-glafs was 

 unable to difcover any jun6lures. This circumftance, and the 

 continuation of the pidure throughout the whole fubftance, 

 rendered it extremely difficult to form any dired notion of the 

 It is found to procefs or manner of performing fuch a work. And the con- 

 confift of threads ^gpjJQj^ of it might have long continued enigmatical, were it 

 endwifc. not that, on the fedion of the fradure mentioned, lines are 



obfervable, of the fame colours which appear on the upper 

 furface, that pervade the whole mafs from one fide to the 

 Other J whence it became a rational conclufion, that this kind 



of 



