176 L. B. WALLACE. 



the loops in the spireme (a and c), and then open out into double 

 V's (//). In the metaphase these double V's are so placed at 

 the equator of the spindle that the longitudidal split is parallel to 

 the spindle axis. This can be determined with certainty since 

 often the splitting occurs after the chromosomes have taken 

 up their position in the equatorial plane and can be seen at 

 different stages of the process, the line of the split being always 

 parallel to the spindle axis (PI. I., Figs. 18, 19, and text figure, e). 



Double spindle fibers connect the ends of the double V-shaped 

 chromosomes with the centrosomes. To express it in another 

 W ay, after a V-shaped chromosome has become a double V by 

 a split passing longitudinally along each arm, each half of each 

 arm is connected by one linin fiber with the centrosome, but 

 since the distal ends of every such pair are in close juxtaposition, 

 we have the appearance of double spindle fibers passing from 

 them to the poles. The first maturation division then takes place 

 through what corresponds to the apex of the original V's and is 

 a transverse or reduction division (/"). When the centrosomes 

 divide to form the daughter centrosomes for the second matura- 

 tion spindle, one of each pair of linin fibers remains attached to 

 one of the daughter centrosomes. When the latter have moved 

 in opposite directions and have taken up their positions at right 

 angles to the position of the mother centrosome (thus becoming 

 the poles of the second maturation spindle) the V-shaped daughter 

 chromosomes, formed from the longitudinal splitting of a single 

 arm, are left attached by their apices in the equator of the spindle. 

 The space between the arms of these single V's corresponds to 

 the space between the halves of the longitudinally split spireme 

 and, therefore, when the arms of the V are drawn to opposite 

 poles, we plainly have an equational division (g, h, i). 



To sum up : The first division occurs at a point corresponding 

 to the bend of the loop in the spireme while the second occurs 

 along the line of the original, longitudinal split of the spireme. 

 Such an interpretation must surely be the true one if it be granted 

 that the loops arise in the synapsis by a conjugation, end to end, 

 of like chromosomes. And, even if the loops are supposed to 

 originate by the spireme simply breaking up into the reduced 

 number of chromosomes, still the sequence of the divisions would 

 be as given above. 



