426 HISTOLOGY 



into qualitatively and quantitatively equal halves is called an equation 

 division. If the division is of the ordinary mitotic type, it is known as 

 homeotypic; if the prophase of the division is characterized by various 

 ring- and cross-shaped chromosomes, the mitosis is said to be heterotypic. 

 If the reducing division (frequently heterotypic) precedes the equation 

 division, the case is known as prereduction; the reverse condition presents 

 a case of postr eduction. Both methods have been observed among ani- 

 mals and plants and with about equal frequency. Indeed, in some 

 cases investigators disagree as to which is the method in the same species 

 of animal. In very many recorded and well-authenticated cases, how- 

 ever, one of the maturation divisions is a reducing division and the other 

 is an equation division. 



In spermatogenesis, where the four cells resulting from the matura- 

 tion divisions of a spermatogonium are all functional, the reducing 

 division effects a qualitative inequality between the two pairs of fission 

 products, as demonstrated by Wilson in certain Hemiptera; this dis- 

 similarity may be in part a sex-determining factor, as McClung first 

 suggested. The reducing division in oogenesis, where only one of the 

 four cells resulting from the maturation process remains functional, 

 effects the loss of chromatin to the ovum that may have represented sex 

 determinants and various other ancestral hereditary characters. Ac- 

 cording to Weismann and his followers, the gist of the maturation phe- 

 nomenon lies in the redistribution of the morphological representatives 

 of hereditary characters, and offers the basis for variation and selection. 



To take up a concrete example, let us consider the maturation proc- 

 esses of the starfish, Asterias forbesii, where the chromosomes are all 

 characteristically dumb-bell-shaped throughout the maturation divi- 

 sions. Here the somatic number of chromosomes is about 36. In syn- 

 apsis, which probably occurs during synizesis, this number is reduced 

 to 18, so that, with possibly one or two exceptions, the resulting com- 

 pound chromosomes are bivalent. These chromosomes are typically 

 bi-lobed. Since the details of synapsis in this particular case have not 

 yet been observed, we have no clew as to what the maturation divisions 

 mean where two longitudinal fissions are known to occur. But let us 

 consider the various possibilities. 



If the chromosomes united end to end in synapsis (telosynapsis) 

 and condensed into a bi-lobed chromosome, so that one lobe is A (the 

 parental chromosome) and the other is B (the maternal chromosome), 

 then the first longitudinal division yields chromosomes that must be 

 represented by AB, and the division was an equation division. The 

 second longitudinal divisions would again result in chromosomes AB, 

 and no true reduction would have taken place. Suppose the chromo- 

 somes to have fused side by side in synapsis (parasynapsis) and condensed 



