38 NAGELl's THEORY. [BOOK r. 



motile spiral threads, spermatic filaments, are formed in Flo- 

 ridese, Hepaticse, Musci, Characese, and Filices. I believe, 

 however, that the cellules in that case have not the signification 

 of cells, but of nucleal vesicles, and therefore that the nuclei- 

 form body is not a nucleus, but a nucleolus. This assumption 

 is perfectly natural in the Characea3, as there the cellules of the 

 spermatic filaments are to be distinguished as special structures 

 inside the actual cells. In other Cryptogamia, which possess 

 the so-called spermatic animalcules, experience can offer proof 

 neither for nor against, as at present we are altogether without 

 facts as to the history of their development. 



" As to the structure of the nucleoli, in my opinion nothing 

 universal or precise can as yet be said. In some cases they 

 .appear to be merely accumulations of mucilage. In others, 

 a membrane which surrounds them is wholly unmistakeable. 

 In every case it is certain that they always appear with a 

 clearly defined margin. This circumstance speaks strongly 

 in favour of the assumption that they, like the nuclei, are 

 inclosed in an utricle. Since, if they were merely agglome- 

 rated mucilage, we should have nucleoli, the substance of 

 which would pass gradually into the mucilage of the nucleal 

 vesicle, and which would generally possess an irregular peri- 

 phery. Or if they originated by the deposition of layers 

 from without, this lamellar structure would be perceptible in 

 large and perfect nucleoli. Not only is there nothing of the 

 kind to be seen, but the facts of the appearance of hollow 

 space, the whole mass becoming vesicular, and the change 

 into a granular substance, testify against it. The presence 

 of a membrane is now indeed indubitable in some nucleoli. 

 It is, however, still a question whether this is not some- 

 what accidental, whether it does not represent merely the 

 thickened outermost layer of mucilage; since also where 

 a hollow space makes its appearance in the contents of 

 the nucleal vesicle, around the nucleolus, this becomes 

 hardened into a membranous structure on its bounding sur- 

 face. So long, too, as the membrane of the nucleolus, 

 from its vital history, does not appear to be anything essen- 

 tial (as that, for instance, in the nucleal vesicles resulting 

 from propagation), I cannot, from its presence in individual 



