RHIZOrUS NIGRICANS. 19 



the ba.se of the sporiiiigiuni thvouoh the denser phifsiii (PI. 11. tig. 8). 

 This furrow increases in depth until it reaches and fuses with the 

 lowest vacuoles in the hiyer. Thus the protoplasm of the sporangium 

 is divided into two distinct portions destined to perform radically 

 different parts in the further life of the plant. That outside the cleft 

 is to be entirely cut up into spores, while that inside is later to be 

 surrounded by the columella wall and plays no direct part in repro- 

 duction. The former I sliall distinguish as the spore-plasm and the 

 latter as the columella-plasm. It will be noted from what has been 

 alread}' said and from PI, II, tigs. T and 8, and PI. Ill, tigs. !<» and 12, 

 that the columella-plasm includes all the looser plasm in the sporan- 

 gium and also a thin layer of the denser plasm. 



Orve might have expected from PI. 1, tig. (!, that the columella 

 wall would l)e laid down in the clear zone shown in that tigure, but 

 that such is not the case there is no room for doul)t. The writer has 

 preparations in which this zone is still almost as marked as in the 

 tigure mentioned, while the columella cleft is forming in the denser 

 plasm. PI. 11, tig. 8, and PI. Ill, tig. 1(>, show that the outer part of 

 the looser plasm is still souunvhat clearer than that in the center, 

 though the paths of the currents have become almost o))literated. The 

 time for the disappearance of the currents varies greatly in ditierent 

 sporangia. 



There is no visi))le ditl'erence while cleavage is going, on between the 

 denser plasm inside the layer of vacuoles and that outside, nor is 

 there any diti'erentiation of the c3'toplasm between the vacuoles or in 

 advance of the surface furrow, such as Harper found in the late sub- 

 divisions of the protoplasm of Piloholus and in the last stages of cleav- 

 age of FuUgo (lOOo). 



While the cutting out of the columella is going on, the sporangium 

 gives every appearance of having only slight turgidity. The cleft in 

 the protoplasm is always quite wide — at least in certain places. When, 

 however, the cleavage is complete, the protoplasmic masses increase 

 in volume and become strongly turgid again, causing the two proto- 

 plasmic surfaces lately separated to become pressed together so tightly 

 that onl}^ 1)}' the closest study can one follow the cleft throughout its 

 entire extent. 



In case the spore cleavage, which will be described later, begins 

 before the columella cleft is completed, as often occurs, this period of 

 turgidity is postponed until after the spores are entirely cut out. 



It will be noted that when tirst formed the cleft around the colu- 

 mella is bounded by two protoplasmic surfaces. When these surfaces 

 become tightly pressed together b}' the turgor in the sporangium, one 

 might expect them to fuse into a continuous mass of protoplasm again, 

 there being no wall between them at this time. Indeed, such a phe- 

 nomenon was described by Biisgen (1882) in the formation of the 



