68 
H. E. Jordan 
(equal?) among the daughter spermatocytes where they are said to undergo 
a numerical increase. (Duesberg earlier. 1907, noted chondriosomes first 
only at the end of the growth period of the primary spermatocyte in the 
guinea-pig.) 
The long list of observations, however, indicating a first appearance 
of mitochondria in the growing spermatocyte (oocvte) cannot be ignored. 
Duesberg rneets this array of contradictory evidence (i. e. to his conti- 
nnity theory) with the Statement, “faulty technique”. Such a Statement 
must, however, further include an explanation of why (as in my own 
material of opossum testis) the “imperfect” fixation and staining reveals 
the chondriosomes only at certain stages and not at others. In the testis 
of the opossum there are possible mitochondrial elements (very similar 
in form and staining quality to true chondriosomes) among the deuto- 
plasmic spheres and granules of the spermatogonia and trophocytes. 
But there is a very conspicuous absence of either elements in the young 
spermatocytes. Only in later stages of the growth period (i. e. post- 
svnaptic stage — pachytenic phase of the nucleus) do they appear in 
quantity; and their appearance is coincident with a transient achromatic 
reticular phase of the nucleus, and an extrusion of nuclear chromatin. 
The extrusion cannot be proved by the cvtological evidence of a transit 
through the nuclear membrane (i. e. lacunae in membrane and granules 
caught half-way) but is inferred from the fact that intra-nuclear and 
extra-nuclear particles have a similar form and staining reaction. Sub- 
sequently half the periphery of the nucleus may be surrounded with 
closelv adhering compact spherical or bilobed ehromidia. This phase is 
coincident with an absence of chromatin particles within the nucleus. 
The appearance of chondriosomes in clumps throughout the cytoplasm 
is coincident with the disappearance of the above elements externally 
on the nuclear membrane. The evidence thus points strongly, I believe, 
to a nuclear (chromidial) origin of chondriosomes. 
Montgomery (1910) has shown clearly in Euchistus that size of 
spermatocyte (and amount of the mitochromidial substance in the cyto- 
plasm — regarded as nuclear in origin) is closelv related to size of Sertoli 
cells (i. e. nutritional factors). The amount of mitochondria seems due 
to the quantity of food-supply — and the metabolic activity of the cells 
is under the control of the nucleus. If mitochondria can increase in 
number with increase of nutrition, they may very well disappear after 
periods of great activity, as after spermatogonial division. Such evanes- 
cence renders the interpretation of mitochondria as hereditary Anlagen, 
as Duesberg suggests, very improbable. Thus they may be absent in 
