rEXOTllEKA I\[IJT. LA'I’A AND (K. :\rU'I'. SK.'\nLATA. 
537 
of cliroinosomes present could be determined with the utmost 
exactitude. Moreover, in each individual the chromosomes 
in a number of nuclei were counted independently by us both, 
the same conclusions being reached in each case. The con- 
stancy in chromosome number is such that a few perfectly 
accurate counts are amply sufficient to determine the number 
present in any individual, though a larger number of counts 
were made in nearly every case. 
In plant No. 226 . II . 18, which was intermediate between 
typical lata and semilata, fourteen counts made in the 
})etals and ovaiy gave 15 chromosomes in every case. In 
addition, six counts of the heterotypic metaphase each gave 
15 chromosomes, and numerous later stages of meiosis 
showed the 7 + 8 distribution of the chromosomes in the 
reduction division, which will be discussed later. No. Ill . 1 
of this culture, a typical lata plant, also showed 15 chromo- 
somes, both in its somatic and meiotic divisions. 
In cult. 229, the nine plants having lata or semilata 
characters all possessed 15 chromosomes. In one of 
these plants (No. I. 4) several cases were observed in which 
the somatic chromosomes at inetaphase were more or less 
completely divided transversely into two (see figs. 12 and 13). 
A light area in which not even linin could be detected was 
found usually in the middle, but sometimes nearer one end of 
each chromosome. A similar condition was observed by 
Fraser and Snell (1911) in Vicia fab a, but only in one or 
two of the chromosomes, whereas in the present case nearly 
all the chromosomes of a group are thus divided. Similarly, 
Agar (1912) has found in Lepidosiren that the chromosomes 
frequently undergo a transverse segmentation which includes 
the chromatin, but not the linin portion of the chromosome. 
This segmentation does not always occur in the middle, but 
sometimes nearer one end, and moreover, the point at which 
constriction or segmentation takes place is found by him to be 
constant for each chromosome. Numerous other cases of 
transverse segmentation are to be found in the literature. 
In a semilata plant (1 . 6) was found the condition shown 
