292 



BOTANICAL GAZETTE 



In these apogamous prothallia antheridia are always produced in 

 profusion and sperms are matured in every case, but the development 

 of archegonia is different in different forms : sometimes no archegonia 

 are produced (var. polydactyla Wills, and cristata apospora), or some 

 are formed but do not attain maturity (var. darissima Jones), and 

 in these two cases sporophytes appeared apogamously as vegetative 

 outgrowths; in other cases an egg is produced in the normal way, 

 but there is no fertilization, the embryo being developed either from 

 the unfertilized egg (var. darissima Bolton and Scolopendrium) or 

 endogenously in connection with an archegonium (unco-glomeratum 

 Stansfield). 



So far as the number of chromosomes is concerned, these seven 

 cases of apospory and apogamy may be placed in three categories: 

 (1) in the two vars. polydactyla, in which sporogenesis shows the ordi- 

 nary reduction, a doubling of chromosomes is attained by the fusion of 

 vegetative nuclei instead of by fertilization; (2) in four forms — three 

 varieties of Athyrium Filix-foemina and a Scolopendrium — sporogene- 

 sis is omitted from the life-cycle, prothallia arise directly from abortive 

 sporangia or pinnae, and sporophytes develop apogamously from the 

 prothallia, the approximate sporophytic number of chromosomes being 

 retained throughout the cycle; (3) in var. cristata apospora, in which 

 apospory and apogamy regularly follow each other, the approximate 

 number of chromosomes through the life-cycle is 60, and the authors 

 believe that in this case the sporophyte may retain the original game- 

 tophytic number of chromosomes, the suggestion being based upon 

 the fact that the number 60 approximates 72, the gametophyte 

 number of the type species Lastrea pseudo-mas. 



Throughout the investigation the authors met a variable number 

 of chromosomes in different parts of an individual, and they express 

 the belief that such variations are not due entirely to errors of counting, 

 but that they correspond to fluctuations in the number actually present 

 in the different regions. Their general conclusions are as follows: 

 there is no necessary correlation between the periodic reduction in the 

 number of chromosomes and the alternation of generations. Fertiliza- 

 tion and reduction, however, are recognized as holding a very definite 

 relation to one another, but without any a priori grounds for assuming 

 any necessary connection between either of them and any other 



