50 ALTERNATING GENERATIONS 



and various Polypodiaceae. In the Bryophytes, the earliest case described 

 was that of Pallavidnia investigated by Farmer, with four chromosomes 

 in the gametophyte and eight in the sporophyte : and the same relation 

 between the generations has since been found also in other genera of 

 Liverworts. In Mosses few observations have yet been made, but Mr. 

 M. Wilson has found that in Miiiuni liornian the numbers are 6 and 1 2 

 respectively in gametophyte and sporophyte. Accordingly the difference 

 in chromosome-number may with high probability be held as a general 

 diagnostic feature between the two generations in normal representatives of 

 the Archegoniate series. 



This being so, the recognised limits between the generations will 

 naturally be expected to be the points of transition from the one chromosome- 

 number to the other. Now it is found that when the sexual fusion of 

 the two nuclei takes place, the subsequent divisions of the fusion-nucleus 

 show the doubled number of chromosomes : therefore the zygote will be 

 the one limit, and this is in accord with the old distinction of the 

 generations dating from the time of Hofmeister. In his practice the 

 other limit was the spore, since this is the actual body separated as an 

 independent germ. But it is found that the actual reduction of the number 

 of chromosomes to one half, that is, to the original pre-sexual number, takes 

 place at the tetrad-division of the spore-mother-cell. This cell divides 

 twice in rapid succession, and the process is well illustrated in the case 

 of the pollen mother-cells of Lilium, in which it has been specially studied 

 (Fig. 32). It starts from a cell with a nucleus having the double number 

 of chromosomes, as shown by its origin. The nucleus first enters the 

 condition of synapsis (Fig. 32. 3, 4), in which a lateral fusion of the 

 chromosomes in pairs, respectively of paternal and maternal origin, is 

 believed to take place : presently a coiled thread frees itself from the 

 tangle of synapsis (Fig. 32. 4, 5), which becomes shorter and thicker, 

 still showing, however, indications of its double nature (Fig. 32. 6, 7), 

 and divides into segments, which are half as many as those of the parent 

 nucleus (Fig. 32. 7, 8) : each individual of the chromosome-pairs then 

 moves apart (Fig. 32. 10, n), one of each pair passing to either pole 

 of the spindle which has meanwhile been formed : as each half is an 

 original chromosome, the number at each pole is one half that of the 

 parent nucleus, and the division is styled the heterotype, or reducing 

 division (Fig. 32. 12). The second division in each of the two nuclei 

 thus formed follows quickly, and is homotypic, that is, each chromosome 

 undergoes longitudinal fission into two, as in a vegetative division (Fig. 32. 

 13, 14, 15). The four nuclei thus constituted have also half the number 

 of chromosomes present in the nucleus of the spore-mother-cell ; but the 

 reduction is actually effected, as has been seen, in the first, or reducing 

 division. Accordingly, Strasburger has recognised the spore-mother-cell, 

 in which the reduction is initiated, as the actual limit between the two 

 generations. But it is the spore itself which normally terminates the 



