FUESH-W ATKlt ALUJi OF THE UNITED STATES. 219 



These arc borne upon cells similar to those supporting the antheridia. At first 

 tin \ arc not markedly different from the other cells, but soon undergo a very rapid 

 growth. This is not, however, regular, and is not partaken of by a band of tissue 

 about one-third way from the basal end, so that at last a long somewhat flask- 

 shaped cell is produced, with a very marked contraction at the point indicated, 

 separating it into two portions. The wall of this cell is thin but very distinct, and 

 the cavity is filled with a homogeneous or very sparsely granular protoplasm, which 

 is continuous through the narrow neck-like portion. After a time there appear 

 one or more large irregular vacuoles, with actively moving corpuscles in them, and 

 at the same time the neck appears to be stopped with a slimy substance. Careful 

 examination with reagents shows that this is cellulose, and that it does not com- 

 pletely block the passage-way through the isthmus. At this time there appear lying 

 upon the free end of the trichogonia globular or flattened bodies, without external 

 membrane, corresponding in all respects with those already described as being pro- 

 duced iu the antheridia. The end of the trichogonium generally enlarges at this 

 period into a sort of roundish knob, and by and by the end wall between this and 

 one of these globules becomes absorbed, so that there is a free communication 

 between the two. Whilst this is going on the globule acquires a thin, delicate 

 coat, and there appears in it a vacuole similar to those preexisting in the tricho- 

 gonium. 



The first result of this impregnation of the trichogonium is the deposit of new 

 cellulose, and the complete blocking up of the passage-way through the isthmus 

 or narrowed portion. Already before the fecundation, the upper cells of the 

 branches supporting the trichogonia have produced numerous branchlets, which 

 Crowing upwards more or less completely cover that organ. After impregnation 

 the cells near to the trichogonium become much larger and broader, their vacuoles 

 disappear, and arc replaced by a dense granular dark greenish-brown protoplasm. 



These cells now show a great activity in the production of numerous branches 

 in the usual way, but it is the upper two alone which, with the trichogonium that 

 they support, are concerned in the formation of the fruit glomerulus. These put 

 out all over their surface an immense number of protrusions, which soon in the 

 ordinary way become the parents of as many twigs or branchlets, which growing 

 and branching, precisely as do the vegetative branches, soon become excessively 

 crowded. The base of the trichogonium participates also in this production of 

 branches, and at last a dense ball is formed of pseudoparenchymatous tissue by 

 the forced adhesion of the crowded twigs. The central cells of the glomerulus 

 thus formed are very large and bladder-like. The outer part of the ball is com- 

 posed of innumerable radiating rows of small cells, the end cell of each branch 

 being roundish so as to present a convex external face. At maturity these 

 cells open and allow their contents to escape as round masses, which appear to 

 have no membrane, but begin at once to grow and secrete cellulose. Their after- 

 hi-tory has not been made out with absolute certainty, but they are believed to 

 directly develop the new plant. 



