CHROMOSOMAL EVOLUTION AND SPECIATION
115
acrocentric chromosomes of the grasshopper Mecostethus what we interpret as the "short arm" is in fact the centromere, but we cannot consider
our interpretation (White, 1935) as disproved by his photographs.
The idea that the natural ends of chromosomes differ from the newly
formed "ends" exposed by artificial or spontaneous breakage is due to
Muller (1932, 1938, 1940a,b) who has used the term telomere
to
designate the natural ends. Telomeres ordinarily show no tendency to
unite with other telomeres or with freshly broken "ends," whereas the
latter have a strong tendency to fuse with one another. Thus it is not
possible for a telomere to be transferred to an interstitial position in
the chromosome, as a result of a rearrangement. Apparently, it is also
not possible, in Drosophüa at any rate, for a freshly broken "end" to
persist. The reason for this is probably that a "sister strand reunion"
will occur at the newly formed end, leading to an endless series of
breakage-fusion-bridge cycles. These principles have recently been
verified by a series of ingenious experiments specifically designed to test
them (Muller and Herskowitz, 1954). Whether they apply strictly in all
organisms is somewhat doubtful. In maize, freshly broken ends can
apparently undergo "healing" in the sporophyte, but not in the gametophyte or the endosperm (McClintock, 1941). And some kind of "healing" must occur in the chromosomes of the nematode worms of the
family Ascaridae, which break up into several or many pieces in the
somatic cells. However, the facts of comparative cytogenetics suggest
that Muller's principle of the noninterconvertibility of telomeres and
interstitial regions applies very generally and perhaps universally.
The fact that chiasmata may "terminalize" in many organisms without
the ends of the homologous chromosomes separating prematurely (i.e., before anaphase) probably depends on special properties of the telomere.
Callan (1949) has suggested that at meiosis telomeres remain effectively
undivided long after the rest of the chromosome has split, thereby preventing separation. The idea is plausible but as yet unproven. If it is true, the
anaphase separation of bivalents with pairs of reciprocal chiasmata ("twostrand doubles") on the same side of the centromere should be easier than
that of bivalents with single chiasmata, diagonal pairs ("three-strand
doubles"), and complementary pairs ("four-strand doubles").
The telomere concept has been largely ignored by many plant cytologists such as Nordenskiöld (1951), Vaarama (1954), and Cámara (1954),
all of whom have postulated simple chromosome breakage as an evolutionary process, particularly in the rushes of the genus Luzula, with nonlocalized centromeres. But there are many large plant genera such as
Vinns, Quercus, and Eucalyptus in which all the species have the same
chromosome number, and many others in which all variations in chromo-
115
acrocentric chromosomes of the grasshopper Mecostethus what we interpret as the "short arm" is in fact the centromere, but we cannot consider
our interpretation (White, 1935) as disproved by his photographs.
The idea that the natural ends of chromosomes differ from the newly
formed "ends" exposed by artificial or spontaneous breakage is due to
Muller (1932, 1938, 1940a,b) who has used the term telomere
to
designate the natural ends. Telomeres ordinarily show no tendency to
unite with other telomeres or with freshly broken "ends," whereas the
latter have a strong tendency to fuse with one another. Thus it is not
possible for a telomere to be transferred to an interstitial position in
the chromosome, as a result of a rearrangement. Apparently, it is also
not possible, in Drosophüa at any rate, for a freshly broken "end" to
persist. The reason for this is probably that a "sister strand reunion"
will occur at the newly formed end, leading to an endless series of
breakage-fusion-bridge cycles. These principles have recently been
verified by a series of ingenious experiments specifically designed to test
them (Muller and Herskowitz, 1954). Whether they apply strictly in all
organisms is somewhat doubtful. In maize, freshly broken ends can
apparently undergo "healing" in the sporophyte, but not in the gametophyte or the endosperm (McClintock, 1941). And some kind of "healing" must occur in the chromosomes of the nematode worms of the
family Ascaridae, which break up into several or many pieces in the
somatic cells. However, the facts of comparative cytogenetics suggest
that Muller's principle of the noninterconvertibility of telomeres and
interstitial regions applies very generally and perhaps universally.
The fact that chiasmata may "terminalize" in many organisms without
the ends of the homologous chromosomes separating prematurely (i.e., before anaphase) probably depends on special properties of the telomere.
Callan (1949) has suggested that at meiosis telomeres remain effectively
undivided long after the rest of the chromosome has split, thereby preventing separation. The idea is plausible but as yet unproven. If it is true, the
anaphase separation of bivalents with pairs of reciprocal chiasmata ("twostrand doubles") on the same side of the centromere should be easier than
that of bivalents with single chiasmata, diagonal pairs ("three-strand
doubles"), and complementary pairs ("four-strand doubles").
The telomere concept has been largely ignored by many plant cytologists such as Nordenskiöld (1951), Vaarama (1954), and Cámara (1954),
all of whom have postulated simple chromosome breakage as an evolutionary process, particularly in the rushes of the genus Luzula, with nonlocalized centromeres. But there are many large plant genera such as
Vinns, Quercus, and Eucalyptus in which all the species have the same
chromosome number, and many others in which all variations in chromo-