162 F. 0. BOWER. 



were used ; in all of these the process of plasmolysis was 

 observed, its main features being the same here as above 

 described for the prothallus. 



Special attention was given to the diaphragms of the petiole 

 of Pontederia (Eichornia) coerulea, which consist of 

 flattened, polygonal, thin-walled cells, in close contact with one 

 another, except at the angles where three or more cells meet ; 

 at these points are intercellular spaces, which act as channels 

 of communication between the cavities above and below the 

 diaphragm (cf. figs. 5, 6). On treating a transverse section, 

 including a diaphragm, with 1 per-cent. salt solution, a slight 

 contraction of the protoplasm takes place. In a very large 

 number of cases it is found that the protoplasm first leaves the 

 wall at those points where two cells are separated from one 

 another by a thin septum, Avhile it still remains in contact with 

 the parts of the wall adjoining the intercellular spaces. This 

 would not be the case, if the connection of the protoplasm 

 with the septa were more close than with the walls adjoining 

 intercellular spaces ; hence it may be inferred that it is not so. 

 In fig. 5, which illustrates this, and which was drawn imme- 

 diately after plasmolysis, there are no strings to be seen run- 

 ning to the cell walls; but when plasmolysis is more complete, 

 and, after the lapse of a short time, numerous strings may here 

 be seen, as in other cases (fig. 6). It is found that strings 

 run both to the septa and to the walls adjoining intercellular 

 spaces, and no distinction in the numbers which run to these 

 diff'erent parts of the wall has been observed. Comparing this 

 observation with the fact that the strings run in as large num- 

 bers to the free walls as to the septa in the prothallus, it is 

 seen that the same inference may be drawn from both cases, 

 viz. that as far as evidence from plasmolysis goes, the con- 

 nection of the protoplasm is just as close with the free walls 

 as with walls separating contiguous cells. 



It should be noted that also in Pontederia the septa 

 dividing contiguous cells have not a perfectly smooth sur- 

 face, though there are no obvious pits in the usual sense of 

 the term. 



