io6 



LECTURE VII. 



formation of the cellulose wall. The connecting fibrillse hereupon become indis- 

 tinct; they sink together to a structureless mass of plasma, which is withdrawn 

 to the wall of the cell, or they fuse to several coarser fibres, or they simply dis- 

 appear in the surrounding protoplasm.' 



The processes of cell-division hitherto described may be considered as typical : 

 but, under certain circumstances, more or less deviating forms occur. 



Passing over isolated cases, we may turn at once to the processes which have 

 hitherto been distinguished as essentially different from those above described, 

 under the name of 'free cell-formation.' Apart from the cases already mentioned 

 in cryptogamic plants, it is especially in the formation of the endosperm in the 

 embryo-sac within the ovule of the Phanerogams that the so-called free cell- 

 formation occurs. The opinion was for a long time held, that in the protoplasm 

 of the embryo-sac a great number of cell-nuclei arise simultaneously and inde- 

 pendently of one another, as it were 

 like crystals out of a mother-liquor, 

 and that around each of these a por- 

 tion of protoplasm collects which then 

 becomes enveloped with a cell-mem- 

 brane. According to Strasburger's pub- 

 lications, however, the numerous cell- 

 nuclei visible in the protoplasm of the 

 embryo-sac, also arise by the division 

 of one originally existing nucleus; and 

 these nuclei then multiply rapidly by 

 bipartition. ' Between the freely mul- 

 tipl}ing nuclei,' says the observer men- 

 tioned, ' connecting fibrillse are visible 

 in the usual manner. In some cases 

 these disappear very quickly, without 

 increasing; in others they grow but 

 imperfectly; in yet others, however, the 

 connecting fibrillae are seen to become considerably increased as in ordinary cell- 

 division, and a cell-plate appears in them. 



'Between the formation of free endosperm and the formation of numerous 

 coherent cells, a distinction no longer exists. If, for instance, numerous spores are 

 to be developed in a sporangium (cf Fig. 98), or numerous oospheres in an 

 oogonium, or, finally, numerous spermatozoids in an antheridium, the nuclei in 

 most cases are seen in the first place multiplying freely by bipartition, and arranging 

 themselves at approximately regular distances; and then separating layers appear, 

 by means of which the mass becomes cut up into as many divisions as there are 

 cell-nuclei. Each nucleus then occupies the middle of a cell. Connecting threads 

 are here not to be observed.' 



The departures from the above described typical form of cell-division, and 

 especially of nuclear division, go still further in many Algae and Fungi. Thus, 

 Schmitz observed in the older cells of the Characece, in which the most various 

 kinds of nuclear division are to be found, that the nucleus breaks up into two 



Fig. io6.-C( 

 the nuclei produced by 

 AgrimoHia eupatoriun 

 irfied.) 



ng development of partition-walls between 

 divisions in the embryo sac of 

 (After Strasburger— very highly mag- 



