M. Mohl on the Structure of Annular Vessels. 21 



to its edges ; but as the latter is constantly the case, we must 

 reject this explanation of the origin of the line of partition. 



In the developed annular vessel, the rings are either en- 

 tirely isolated, or two or three are joined together in different 

 ways. It not unfrequently happens that the line of partition 

 does not divide the ring throughout the whole of its circum- 

 ference, but that the two superposed rings are united for a 

 space variable in extent ; in which case the parts separated 

 are removed to a greater or less distance from each other, and 

 are placed obliquely to the axis of the vessel. (Plate I. fig. 6, 

 Commelina tuberosa : the same form is often met with in the 

 Canna indica.) 



In other cases, and this is the habitual organization, the 

 rings are removed to a greater or less distance from each other, 

 and are separated by a regular spiral fibre, which, according to 

 the distance of the rings, describes one or more volutions, 

 and frequently even a great number. Of this there are several 

 modifications : very generally from a ring will proceed a spi- 

 ral fibre of the same width as the annular fibre, the distance 

 of whose coils is nearly equal to that of the rings in the 

 portion of the vessel which exhibits this structure (fig. 9, of 

 the Gourd) ; the other extremity of the fibre being similarly 

 annexed to a complete ring, followed by rings, either isolated 

 or again reunited by spiral fibres. 



Very frequently also the spiral fibre placed between two 

 rings does not proceed to a junction with the rings, but its 

 extremities become attenuated and terminate at some distance 

 from the ring. In the stem of the Gourd this is nearly as 

 frequent as the preceding case (fig. 2 a y Commelina tuberosa ; 

 fig. 9, Gourd). 



Often also, from two diametrically opposite points of a ring 

 proceed two fibres in a continuous parallel direction. 



Cases are sometimes met with, although rarely, where two 

 rings are united by fibres slenderer than the annular fibre, 

 which generally form a single coil, or at least only a small 

 number of coils (fig. 1, 7? 8, Commelina tuberosa). This 

 occurs in a very evident manner in the vessels whose rings 

 are not homogeneous, but where the spiral fibre is divided 

 by several fissures into threads united in net-work, as in 

 the vessel represented in figure 1. The width of the fibres 

 uniting the different rings presents no exact proportion to the 

 width of the annular fibre, being sometimes about the half of 

 it (fig. 8), sometimes considerably less (fig. 1). The point of 

 union of the spiral fibre with the annular fibre is especially 

 deserving of consideration. When examined with a sufficient 

 magnifying power, we sometimes find (fig. 7> 8) that a part 

 of the annular fibre separates itself to ascend in a spiral di- 



