THE SPERMATaiENESIS OF DOMESTIC MAilMALS. 237 



of coi-tiiiu cockroaches and c.uuc to tlu; couchisiou tluit it is duo to some 

 iutoriictiou between the chromatin s])iiviii(>s unci the centrosomes wliich gives 

 rise to this phem)iaenun. BrcHNKU ('10) more fully demoastnital the rehition 

 of the ceuti-csome und other substjincea of the ci>ll (nnclenr fis well ns cvto- 

 phusiuic subshiuce), where he says ; " Dug ömtriol vermag Kiir^xT in Kern 

 nnd iin rinsmn nnznzieheu imd vermag ferner die Korumembran in seiner 

 Nachbarschaft aufziiUisen." 



The synapsis: — As already descriljed, in the mouse there Ls no inchcation of 



parasynajsis in this stage, while in the rabbit parallel aiTangoment of two univalent 



spiremes is seen to occur, which exactly corresp:)uds with tliat of the horse. 



In the rat ALLEN ('18) lias shown the paras^'Uiipsis, but ho did not 



imhcate the parallel airangemont of the spiremes in this stage. 



Only from olservation of the spiremes in this stage have we rerx-hal the 

 •view that the methixl of synapsis in the mouse is different fi'om that in the 

 former the conjugation probably taking place by telo3ynr433is but in the latter 

 by i^arasynapsis. 



Nevertheless, in order to completely consider the cpiestiou of the synapsis 

 there are two significant fiicts ^\■luch demand careful oljservatiou. The one is 

 the transformation of chromatin spiremes in the post-syuaptic stage and the 

 other is the construction of chromosomes in the lato firophase. In the post- 

 synaptic stage the longitudinal splitting of the chromatin spiremes can usually 

 be seen in the rabbit luid also in the mouse. Whether the longitudinal 

 splitting of the clu'omatin spireme is entirely the same as that of the somatic 

 chromosomes, or whether this sphtting Ls the result of the conjugation uf the 

 univalent chromosomes, is tlie most significant point in the determination of the 

 method of the sjoiapsLs. The former view was maintained by Goldschmidt 

 ('08), by FiCK ('03), Duesberg ('08 a), Bcchser ('09) and by Jordax ('12), 

 whereas WLlsox ('12) holds the latter view. By careful oljservations in the 

 mouse, as ah-eady descrilied, it is found that the ring-shaped chromosomes are 

 formed by the union of tlie two ends of the conjugated univalent spiremes. 

 Soon after the rings are closed up and the metapliase clu'omosomes ap^x^ar. 

 The longitudinal clefb of the chromosome wdiich appear in the early metaphase 

 (Fig. 36) must therefore I« lookeil upon as a conjugated plane of the two 

 univalent chromosomes. 



