22 Papers from the Marine Biological Laboratory at Tortugas. 
tigations with which I am acquainted it has been reported that the longi- 
tudinal cleavage is first to be made evident in the prophase”’ believes that 
“it is only logical to conclude that this division is completed by the first 
spermatocyte mitosis.’”’ Upon the question of the sequence of the two sper- 
matocyte divisions in Arthropoda the various workers are about equally 
divided. McClung, in the case of some Orthoptera studied, believes that 
the equation division comes first. Von Rath (1895), Henking (1890), 
Paulmier (1899), and Montgomery (1898, 1900, and Igor) arrived at the 
opposite conclusion in the case of various arthropods. Since the same 
result is obtained for the spermatozoa in either event, this point is really 
not of the vital importance it was formerly believed to possess. Neverthe- 
less, it will be noted that during the later prophase, when the longitudinal 
and transverse splits again close up and compact chromosomes are formed, 
the long axis of these chromosomes does not appear to become the short 
axis and vice versa, but the earlier proportions and relations are approxi- 
mately adhered to; therefore, since the first maturation division takes place 
transversely to the long axis of the chromosome, it appears that the first 
division of this mitosis is along the line of the second split marked out in 
the early prophases. If the generally accepted interpretation of synapsis 
is correct, 7. e., that it represents an end-to-end (telosynapsis) union of two 
chromosomes (paternal and maternal, Montgomery, I901), and, further- 
more, if these chromosomes do indeed separate during maturation along the 
line of their previous fusion, then the tetrad figures of the early prophase 
represent bivalent chromosomes divided first longitudinally or in the plane 
of the long axis (fig. 67), followed by a transverse split, 7. e.. along the 
line of previous fusion and perpendicular to the long axis. Accordingly, 
the first division of the long chromosomes at metaphase must represent the 
transverse or second split of the early prophase and is a reducing division, 
since it separates whole chromosomes. Figure 119 shows several of the 
characteristic dividing figures (tetrads) of metakinesis. In the light of 
these figures and the probable relations of the splits here shown to those 
of the earlier prophase chromosomes, I believe that the first maturation 
division is transverse and reductional and the second is longitudinal and 
equational (including also the accessory). De Sinéty interprets both divi- 
sions as longitudinal in Leptinia and Menexenus. Due to the persistence 
of connecting linin threads, Stevens (1905) was able to demonstrate very 
conclusively that in Stenopelmatus (California sand-cricket) the first divi- 
sion is longitudinal and equational. 
The chromosomes vary considerably in size and shape. Figure 80 shows 
a typical spindle with variously shaped chromosomes. Spindles at meta- 
phase invariably show two pairs of reversed 1-shaped chromosomes, three 
pairs of long rod-shaped chromosomes, and the remainder are of the short 
rod or dumb-bell-shaped types. Correspondences of size between the biva- 
