ORTHOPTERAN SPERMATOGENESIS 677 



of the chromosomes separate. The two conditions are shown 

 in figure 64 a, being a drawing of chromosomes from the same cell, 

 and in figure 66. Occasionally rings are formed when the fiber 

 insertion is not median, and in that event divarication of the 

 chromatids takes place only on the side of the larger members 

 (fig. 65, next the E-chromosome) or it may be that the opposite 

 result is reached (fig. 64 g). More commonly however, when 

 the fiber attachment is subterminal, rings are not formed and 

 synapsis occurs between the longer limbs only. This is commonly 

 true of the larger pair of speraiatogonial chromosomes and the 

 result is the E-shaped chromosome or some modification of it. 

 In this element it not infrequently happens that the divergent 

 ends, at the point of fusion, become much extended in the equato- 

 rial plate at the expense of the arms attached to the fibers, and 

 swing round in the form of a ring. This part when viewed later- 

 ally appears as a much extended middle bar of the E, but when 

 seen from above exhibits the ring form modified in appearance 

 by the bent arms above and below. 



All the evidence indicates that the composition of the tetrads 

 is the same in all of the Acrididae studied. There are the same 

 elements of the tetrads to be distributed in each case. It is there- 

 fore apparent that the method of this distribution must differ 

 in Stenobothrus and other similar forms, with rings extended in 

 the axis of the spindle, from that in such as have rings lying in 

 the equatorial plate. A comparison of the annular chromosomes 

 of these two types is most suggestive and seems to offer unques- 

 tionable evidence of the distinction between chromosomes that 

 suffer a reduction division in ,the first spermatocyte and those 

 that are divided equationally. In considering the work of dif- 

 ferent investigators upon the question it will be necessary to enter 

 into this comparison more fully, but in order to give a basis for 

 this I will here briefly note what appear to be essential elements 

 involved in the discussion. 



(1) The composition of the rings is the same in all cases and 

 the point of fiber attachment is fixed for each chromosome. 



(2) Rings that divide longitudinally in the first spermatocyte 

 lie extended in the equatorial plate and have the fiber attachment 



