798 Kemp.—On the Question of the Occurrence of 
They are found in tissue known to contain nuclei with double the 
normal number of chromosomes, which nuclei moreover finally disappear. 
They are singularly like the heterotype reduction-figures of sporo- 
genous cells, showing the X, Y, and tetrad shapes characteristic of the latter. 
They do not, obviously, occur in normal tissue. 
On the other hand, the evidence against their true reduction nature is 
as follows:— 
They occur in the pea alone of a considerable variety of plants, and it 
is significant that the normal pea has chromosomes shorter and thicker in 
shape than are those of Galtonia and the bean, &c. The pea chromosomes 
would consequently show any modification of shape due, for instance, to an 
alteration in their chemico-physical state, more conspicuously than would 
those of the latter plants ; should such an alteration take the form of 
swelling and massing together of their chromatin, a very slight change in 
that direction would give them the stumpy, peculiar appearance of hetero¬ 
typical figures. The normal pea has further a curious break at one end of 
its longitudinally split chromosomes (Fig. 22), and the latter therefore 
present a shape which, with the above modification, might even resemble 
an X. 
Again, the chromosomes in question are more strikingly heterotypical 
in character soon after poisoning, than they are after a longer period of 
subsequent growth. At the same time tetraploid cells are still numerous 
at the later stages, and therefore there would appear to be as much cause 
for hererotypical reduction then as earlier. This seems an indication that 
their shape is due to a chemical or physical change of variable amount ; an 
inference which is borne out by the fact that different strengths of poison 
produce different degrees of alteration in shape. 
Again, their occurrence in the pea alone would lead to the supposition 
that if their function is to reduce the amount of chromatin in the tetraploid 
cells, there would naturally be, in these roots, fewer or none of the multi¬ 
polar and degeneration figures found in those of the other plants examined. 
In the pea these methods would be superfluous in presence of the more 
perfect process of automatic reduction. Actually, however, there are as 
many multipolar divisions in this plant as in the bean, and the degeneration 
figures are particularly numerous. 
Again, ‘ heterotype ’ figures do occur, though very rarely, and not in their 
extreme form, in normal tissue; and then generally in the external cells, and 
particularly in roots fixed in acetic-alcohol, a fixative of which the action is 
certainly more violent than is that of Flemming (Fig. 24). 
Further, it has been impossible to show clearly a double cleavage of 
the chromosomes, such as would be present if they were really hetero¬ 
typical in nature. All instances found of ‘tetrad’ figures were open to 
the interpretation of being ‘ dyads ’ doubled on themselves, and so pre- 
