158 



In their structure, they resemble ordinary spiral vessels in several 

 points ; having, more frequently than otherwise, two or more fibres, 

 forming the same screw, coiled in the interior of the vessel, as in 

 compound spiral vessels : — a " double-threaded" screw having been 

 found in the gourd by Mr. Hassall, a four-threaded one in Urania, 

 and an eight-threaded one in Carina bicolor. 



In addition to this spiral arrangement, there are longitudinal depo- 

 sits of fibres, the number of which varies considerably ; some vessels 

 not possessing more than six or eight, whilst in the larger, more 

 than double the number can be detected. When force is applied to 

 these vessels, they are reduced to a spiral ribbon, and the longitudi- 

 nal fibres break at the same point as the membrane, the broken edges 

 of the latter projecting beyond the edge of the spiral: (PI. xix. fig. 4). 

 The termination is very pointed, and one vessel overlies the other for 

 some considerable space. 



Such is the ordinary structure of these vessels : and I was induced 

 to believe them to be perfect of their kind, and that they presented 

 higher marks of development than the true spiral vessels ; but sub- 

 sequent observations have led me to conclude, that this longitudinal 

 deposit of the elements of fibre is only a state preparatory to the 

 complete developement of a vessel, exhibiting oval or quadrangular 

 dots on its parietes. In examining the vessels as they occur in bun- 

 dles, in Canna bicolor, where annular and spiral vessels (simple 

 and compound) are to be found, it can be observed that all these ma- 

 nifest a tendency to produce longitudinal fibres, but it is only in the 

 compound spiral, that the ultimate conversion of at first a spiral into 

 a vessel with dots, regularly arranged in longitudinal rows, occurs. 

 It is known, from the investigations of Mohl and Schleiden, that the 

 spiral is the type of every other form of vessel ; and vessels, whatever 

 phases they ultimately assume, are, in early life, developed on the 

 spiral plan ; (perhaps an exception occurs in the vessels of woody 

 exogens). The annular can be shown to be formed from a spiral, 

 by two or more turns of the fibre, adhering in various parts of the 

 membranous cylinder in which it is contained ; and ultimately, the 

 connecting portion of the fibre becoming absorbed. 



It can be observed in the reticulated vessel, that the fibres, originally 

 spiral, can be so connected by longitudinal, or longitudinally oblique 

 processes of fibres, as to leave between them portions of the mem- 

 branous cylinder, constituting what is known as a " dot ;" these dots 

 occurring in no regular order. 



If this principle of the formation of dots be carried farther, it is not 



