80 DUCTS MOHI/S VIEWS. [BOOK i. 



ranean stem of Equisetum arvense may be used with much 

 advantage. 



If the very youngest inte modes of the buds of the first- 

 mentioned plant be examined, a single extremely delicate 

 and densely-wound spiral vessel is found in each young 

 woody bundle. In older internodes the spires of this vessel 

 are more distant from each other, and near it may be observed 

 an exterior newly-formed narrow-threaded spiral vessel. But 

 if, at this period, the first-formed vessel is examined more 

 attentively, it will be found that all convolutions are not 

 separated in the same manner from one another, but that 

 almost in regular alternation two entire coils adhere firmly, 

 and one convolution is drawn out. In still older internodes 

 the extension is found to be so far advanced, that the free 

 coil loosened from the cell-membrane frequently stretches 

 in the form of a mere band, with a steep ascent from one 

 ring formed of two closed convolutions to another. When 

 the vessels are still further developed, this elongated coil is 

 seen to be corroded by the reabsorbing action of the cell, and 

 all stages of transition are frequently found present in a single 

 duct. Finally, in still older ducts, the connecting spire is 

 perfectly dissolved ; and there may be observed on the sepa- 

 rate rings the extremities of the original spiral fibre. Schlei- 

 den adds that exactly the same process may be remarked in 

 the subterranean rhizomes of Equisetum arvense. 



To these statements Mohl replies thus, (see his paper 

 translated, with a plate, in the Annals of Natural History, 

 vol. viii. p. 20) : " When we examine the fibre of the perfectly 

 developed annular vessel (for which researches I have been 

 accustomed to use the Commelina tuberosa), we find its orga- 

 nisation absolutely analogous to that of the spiral fibre, in the 

 rings being composed sometimes of an apparently homogene- 

 ous substance, and sometimes exhibiting traces of a determi- 

 nate structure. In the broad fibres of C. tuberosa, the fibre 

 frequently exhibits a great number of shallow linear furrows 

 or perfect fissures, forming a net-work of very narrow and 

 elongated meshes. More frequently still these fissures are 

 found in an uninterrupted line in the medial line of the fibre, 

 or they become confluent, and thus divide the ring into two 



