294 
W. E. AGAR. 
marked when tlie chromosomes are sliortest. This has been 
alluded to several times already, and for definite data I refer 
to fig. 15 a, and to the description of the development of the 
transverse constrictions in the shortening chromosomes of 
the meiotic prophase in my former paper. 
If the chromosomes always segmented in the middle, or, 
failing that, at no definite spot, the matter would possess 
but little interest. But the fact that the segmentation always 
occurs in the same chromosome at the same spot demon- 
strates that the chromosomes possess a constant 
lengthwise differentiation. For if a chromosome were 
liotnogeneous, or if its internal differentiation were not con- 
stant, a transverse constriction developed by physical or 
other means would either always occur in the middle or 
else at no constant spot. 
The theoretical importance of inductive evidence of this 
differentiation needs no emphasis. AVe learn, of course, very 
little about the nature of the differentiation. It may consist 
only in a pliysical or chemical difference between the chro- 
matin at the two ends of the chromosome; or it may be that, 
as demanded by theory, the differentiation lies in the chromo- 
meres of which the prophase chromosome is composed, and 
that when the tug comes the chromosome always gives way 
between that pair of chromomeres (always situated in the same 
spot in the chromosome) which are least firmly attached to 
one another. An indefinite number of other possibilities 
})resent themselves, but all demand the hypothesis that a 
given chromosome is always composed of the same differ- 
entiated portions arranged end-to-end in the same order 
along its length. 
The fact that in the diakinetic pairing of asymmetrically 
divided chromosomes to form the definite tetrads, like ends 
are always applied to like, is another indication of a constant 
differentiation, which is only made apparent by the probably 
physical factor, which produces the visible segmentation. 
Moreover it implies that there exists an attraction between 
like parts of homologous chromosomes, not only between such 
