MacMillan: OfJSERVATioxs ON Nereocvstis 



289 



occur between sieve -tubes and the formative layer of the cortex 



r 



from which they arise, but I have seen no true anastomoses be- 

 tw^een the sieve-tubes and the hyphae of the pith-web, nor have I 

 observed any branching of the sieve-tubes by the formation of 

 lateral emergences such as are so common in the trumpet-hyphae. 

 The great abundance of the trumpet-hyphae and their peripheral 

 position, as well as central, in the pith-web may easily have giv^en 



rise to a misconception. 



ofth 



I 

 I 



Sections of apneumatocyst 



Here the cells become tabellar in form with their long axes 



wall taken from a plant 12 dm. in length in which the pneumato- 

 cyst was approximately spherical and 4 cm. in diameter showed the 

 wall of the cyst to be 4.5 mm. thick. The epidermis is not essen- 

 tially different from other epidermal tissue and in the subepidermal 

 regions a cambial layer is present by divisions of which the wall is 

 thickened. The older cortical cells throughout the pneumatocyst 

 wall have their long diameters parallel with the radii of the organ, 

 until an inner cambial zone is reached close to the cavity of the 



cyst. 



parallel with the surface. The pith-web is altogether destroyecd, 

 but in this material characteristic elongated sieve tubes with narrow 

 lumina are present. The presence of such an inner cambial zone 

 seems to be peculiar to the pneumatocyst area and in this zone 



F 



numerous transverse as well as concentric w:alls are constantly 



The intermediate area between the inner and outer 

 cambium, as the pneumatocyst matures into the retort-shaped 

 body, finally comes to consist of cells elongated in the axis of 

 growth rather than as at first perpendicular ^to this axis. 



////' vonur Lamina. — The basal structure of the 



being formed. 



Structure of the young Lamina.— 

 lamina is identical with that of the stipe. Like the latter it must be 

 considered to consist of central cylinder — in this case a very thin plate 

 of cells — of cortex and of epidermis. My serial sections through 

 stipe, pneumatocyst and young lamina show the progressive change 

 from the cylindrical through the oval and elongated-oval to the 

 thin ribbon-shaped section and enable the homologies of the vari- 

 ous areas of the lamina and stipe to be exactly determined. Cross 

 sections through the lamina of a plant iS mm. in height showed 

 the thickness of the lamina to be a little over .1 mm. while from 

 edee to edge the lamina measured 4 mm. The most marked dif- 



