200 D. H. SCOTT. 



walls lie iu various directions^ a fact which gives an additional 

 appearance of complexity to parts of the mature tissue. Where 

 a vessel is iu contact with parenchyma, short pointed protru- 

 sions are very often wedged in between the cells of the latter. 

 But these protrusions never grow out into branches of any 

 considerable length, like those so often described in the case of 

 the Cichoriacesej &c. This fact is of some interest, as the 

 occurrence of these protruding branches is the only certain 

 point in which the laticiferous vessels at all approach the lati- 

 ciferous cells, and it is remarkable that here, where indications 

 of a transition might perhaps have been expected, they should 

 be absent. 



The hypodermal system of laticiferous tubes agrees in the 

 main points of structure with that in the outermost regions of 

 the phloem. The main trunks of the vessels, "which are here on 

 the whole further apart, are connected into a network by 

 branches running more or less accurately in the transverse 

 direction (fig. 1) . The development of the tissue in this region 

 is altogether more scanty than in the phloem, as stated above. 

 I have not especially investigated the mode of origin of this 

 portion of the tissue, but the reticulate arrangement leaves 

 little doubt that it is the same as in that of the phloem. 



Considering the great development of the laticiferous tissue 

 in Manihot, a corresponding reduction of the sieve-tubes would 

 be expected. This at least is the conclusion to which one 

 would be led by the prevailing views as to the relation between 

 the two tissues^ as expressed, for example, by De Bary, especially 

 at p. 541 of his ' Vergleichende Anatomic.' As a matter of 

 fact, however, the reverse is the case in Manihot^ as the sieve- 

 tubes are remarkably well-developed, both in the primary and 

 secondary bast. In the former they have the horizontal or 

 slightly inclined walls usual in this region, each with one sieve- 

 plate. They usually attain a diameter of 0-026 mm. In the 

 secondary phloem their terminal walls are inclined to the radial 

 plane at an acute angle, and each wall is occupied by a single 

 series of numerous sieve-plates. The plates show the connect- 

 ing strands of protoplasm, callous deposits, &c., remarkably 



