72 



JOURNAL OF THE 



[May, 



is advisable to draw them out not more than half an inch, or 

 their strands will be more or less separated, and will resemble 

 in appearance straightened threads. 



A thin longitudinal slice of the stalk shows that the spiral 

 vessels, even when lying in immediate contact, vary greatly in 

 size, from xijWth to xinrffths of an inch in diameter; and the 

 fibres themselves vary correspondingly, the largest being about 

 i^iRnrth of an inch in diameter. 



One of the distinguishing beauties of the larger ribbons is 

 occasioned by the fact that their fibres are crossed nearly at 

 right angles by another set of apparent fibres of extreme fine- 

 ness ; this second set divides the surface of the ribbon into 

 minute spaces, which are nearly squares or parallelograms, 

 according to the distance between the cross lines, and gives the 

 appearance of lattice work constructed of resplendent glass 

 rods. 



Spiral fibre of the Banana stalk. — A portion of a large spiral vessel. 



The longitudinal slice shows that the second sets of fine fibres 

 run longitudinally through the entire length of the large spiral 

 vessels. They do not appear to occur in the smallest vessels 

 but are most numerous in the largest vessels, where a half longi- 

 tudinal section will show as many as ten, and they gradually de- 

 crease in number in the respective vessels of decreasing size, 

 until finally only two or three can be faintly discerned. These 

 sets of fine longitudinal lines, crossing nearly at right angles the 

 larger spiral fibres, appear to be the outlines of delicate, length- 

 ened, nearly rectangular cells ; and these cells appear either to 



