206 D. n. SCOTT. 



In the outer (primary) portion of the phloem of the stem^ 

 the tubes occur either singly, as is most frequent, or two side 

 by side. Where the latter is the case, the intermediate longi- 

 tudinal wall shows frequent perforations. The main trunks 

 are connected into a very regular reticulate system by nume- 

 rous transverse branches, passing between the parenchymatous 

 cells. 



In the secondary cortex the laticiferous tubes occur two or 

 more together, forming bands which enclose the medullary 

 rays. At the upper and lower ends of the latter the bands 

 join. The tubes are connected among themselves by extremely 

 numerous perforations in the lateral walls. These perforations 

 are more numerous even than in Manihot, and it is quite usual 

 for only short fragments of the wall to be left between them. 

 These fragments appear swollen. Where several tubes meet 

 at the points of junction of the meshes of the network, they 

 are all in the freest possible communication one with another 

 by means of these perforations. 



The laticiferous tissue of the outer cortex is very sparse 

 compared to that within the bundle-sheath, and hence anasto- 

 moses are comparatively infrequent. But they occur here as 

 well as in the phloem, and where two or three tubes run side 

 by side, as sometimes happen, they anastomose in a complex 

 manner, the intermediate walls showing numerous perforations 

 of various dimensions. 



On observing the young laticiferous tissue in the neighbour- 

 hood of the cambial layer, I found that the cells from which it 

 is derived abut on one another with very oblique walls (inclined 

 to the radial plane). These walls do not become wholly 

 absorbed, but break down at numerous points. It is not 

 improbable that transverse septa are also present in the earliest 

 stages of development. 



Although my observations on the development are thus by 

 no means complete, yet those which I have already made, 

 taken in connection with the whole structure of the mature 

 tissue, leave no doubt at all on my mind that we have here 

 once more to do with laticiferous vessels, and not with latici- 



