The Spermatogenesis of the Opossum (Didelphys virginiana) etc. 65 
ander the influence (if not as parts) of the nucleus. Duesberg has drawn 
the hne sharply and states the matter unqualifiedly: mitochondria either 
arise in the cell during the process of growth or they have always been 
there. Duesberg does not admit that Goldschmidt has demonstrated 
the nuclear origin of his “chromidial net”, nor thinks that he has establi- 
shed the identity between that body and mitochondria. He points out, 
moreover, that in his recent work (1909) Goldschmidt has yielded the 
following important points: 1) that the chondriosomes may react to 
certain stains differently from chromatin; 2) chromidial-formation is not 
limited to the synapsis phase; he now recognizes chromidia in spermato- 
gonia; 3) ability of chromidia to multiply in the cytoplasm; 4) recognizes 
the correctness of. Meves’ and Duesberg’s observations with respect 
to the formation of neuro- and myofibrils. 
The difference becomes narrowed down to a question of first 
origin, viz. : Do chromidia arise from chromatin material of the nucleus, 
or are they integral parts of the cytoplasm transmitted from cell to cell 
and from parent to offspring? Vejdovsky (1907) has interpreted the 
chromidial-nets described by Goldschmidt in the somatic cells of Ascaris 
as fixation artifacts, i. e. greatly altered and torn threads of the normally 
filar framework of Ascaris cells. Bilek (1909) has still further developed 
this idea and illustrates appearances in the cells produced by violent 
f xing reagents very similar to the chromidial-nets of Goldschmidt. In 
view of the widespread appcarance of this structure in well fixed material 
of very different origin, however, the results of Vejdovsky and Bil'ik 
have only a very superficial value and only a historical interest in this 
connection. 
The observations of many workers upon chondriosomes of the germ- 
cells of vertebrates have shown that the granulär form of the mitochondria 
seems to be the rule among mammals. In lower vertebrates (e. g. amphibia) 
and invertebrates (e. g. Euchistus ) they 'may be füamentous. Moreover, 
in the spermatids of mammals the chondriosomes do not condense into 
a Xebenkern (as in insects) but remain separate until they fuse into the 
spiral filament. 
In the cliick Duesberg (1909) has described the multiplication of 
chondriosomes by transverse contriction (fission). Similar division phe- 
nomena have been described by Gigi.io-Tos and Granata (1908) for 
Pamphagus , and by Faure-Fremiet for a number of protozoa. The latter 
investigator also cannot accept the hypothesis of a nuclear origin in the 
protozoa studied. Apropos staining quahty, he reports that the mito- 
chondria of protozoa ( Strolilidium gijrans ) may be colored by osmic 
Archiv f. Zellforschung. VII. 5 
