( 195 ) 



thin and hollow filaments I have mentioned, the hollows or cavities 

 in which I have caulbd to be reprefented by a fort of points or do's. 

 The biack fliade in this figure, denotes minute veflels furrounding the 

 others, which by reafon of their exceeding fmallnefs could not be 

 reprefented in the drawing. 



Thefc veficles, which I have mentioned to adhere to the cap'lla- 

 ments, as they appear, when cut by a firaight fection, are Ihewn at 

 Jig. 4, EFG H. In thefe veficles there appeared fome fmall parti- 

 cles, concerning which, however, I could not pronounce any thing 

 with certainty. 



After I had made the preceding obfervations, I received from a 

 friend who obferved how defirous I was to inveftigate the nature of 

 the Cocoa wood, a piece of that wood which he had procured from 

 the ifland of Curacoa. This was feven inches in diameter, and about 

 four inches long : in the middle, it was compofed of the before men- 

 tioned larger capillary parts, but, on the outfide, and about an inch 

 from the furface, it was fo hard, that in attempting to fplitit, I broke 

 a fieel wedge in pieces, and I do not remember ever to have met 

 with fo hard a wood. 



Between the bark and the folid part of this wood, I law fome 

 capillary parts creeping along, and which v,'ere of the fame nature 

 as thofe pi6lured mjig. 1, between G and N, and thefe capillaments, 

 I was informed, are made ufe of to be twifl;ed hito ropes and ca- 

 bles. 



I obferved, that wherever the knots or bands in the bark ap- 

 proached each other, as at F K, there the capillaments grew out of 

 the bark : fometimes I obferved one or two branches rife out of one 

 of thefe capillaments, and thefe again fubdivide into I'maller ones, 

 hollow within, and which capillaments I judged might in time unite 

 and form velfels of the fize pictured mfg. 2, at T T. And if fo, it 

 follows that the Cocoa tree receives the addition to its bull< from ther 



