Multinucleate Cells : An Historical Study (1879-1919). 29 



conditions. The main difference between our results and those of 

 the other authors quoted is tliat, according to our observations, the 

 plurality of nuclei arises, not by aniitosis, but by a process of 

 karyokinesis with which certain peculiar features are associated. 

 In 1919 we published a fuller paper * in which we dealt on broad 

 lines with the occurrence of multinucleate cells in vegetative 

 tissues. We recorded the occurrence of more than one nucleus — 

 two being the commonest number — in the young parenchymatous 

 tissues of 177 species representing 60 families, including members 

 of the Filicales, Equisetales, Lycopodiales, Psilotales, Isoetales, 

 Gymnosperms, Monocotyledons and Dicotyledons. Our observa- 

 tions related chiefly to stems, but w^e also found a binucleate phase 

 in leaves and roots. It is most conspicuously developed in 

 parenchymatous tissues, such as the cortex and pith of the axis, 

 and the mesophyll of the leaf, but we have also seen it in the 

 central cylinder. The " heads " of Asparagus, at the stage at which 

 it is usually cut, form particularly favourable material on which to 

 demonstrate the binucleate phase, which is a striking feature of the 

 ground-tissue cells. 



As we have already meiitioned, we find that the binucleate 

 condition is invariably brought about by mitosis. The division 

 occurs normally in the earlier stages, up to the period at which 

 the two daughter nuclei are at the poles of the spindle, while the 

 cell-plate is just being initiated. But at this point the mechanism 

 seems to break down and the cell-plate is resorbed, while the 

 phragmoplast,t with its associated cytoplasm, goes through a 

 singular metamorphosis. It becomes vacuolate in the centre and 

 develops into a hollow sphere which gradually grows until it 

 encloses both the daughter nuclei, and then, by its further extension, 

 ultimately merges into the cytoplasm lining the cell wall. For 

 this hollow shell we have proposed the term " phragmosphere." 

 In some cases it is exceedingly well-defined and stains deeply, 

 giving the sections in which it occurs a curious appearance of 

 exhibiting cells within cells. 



The binucleate condition of parenchymatous cells persists in 

 some cases for a very long time — possibly throughout the life of the 

 -element — but in other cases the cells eventually become uninucleate. 

 This seems to be brought about by the degradation and resorption 

 of one nucleus of the pair. We have seen no evidence of any 

 other method of transition from the binucleate to the uninucleate 

 state — such, for instance, as fusion of the nuclei, or a belated 

 development of walls between them. In old tissues, lobed nuclei 

 are frequently seen which might easily be taken to be stages either 



* Beer, R. and Arber, A. (1919) ; see also Arber, A. (1920). 

 t This convenient term wag introduced by Errera, L. (1888), to denote the com- 

 plex ofspindle fibres which generally assumes the form of a ** Rotationsellipsoid." 



