APOGAMOUS FERNS. EVOLUTION OF THE SEPARATE SPECIES 



the roof of the University in which no other sporing ferns were present. Next spring, 

 when new leaves had expanded, uncontaminated spores were collected and sown on to 

 soil after sterilization of soil and pot in an autoclave. During culture the pots were 

 kept covered with a glass lid and were watered only from below by standing them in a 

 zinc tray; a chance admixture of stray spores from outside seemed therefore excluded. 

 On germination, the usual sparse crop of apogamous prothalli was produced which 

 were removed to another pan as soon as they appeared. The residue was then closely 

 scrutinized. A few abortive apogamies were found, as described by de Bary, but, in 

 addition to these, five specimens were found with definite and fairly abundant arche- 

 gonia in the usual position. All these five looked more or less abnormal, recalhng indeed 

 the type of aberrant and depauperate prothalli which result from a sowing of spores 

 from triploid Osmunda. Attempts to induce the formation of sexually produced young 

 plants on them were unsuccessful and all died without offspring, although one remained 

 alive for two years making very slow growth before it succumbed. None showed the 

 slightest attempt at apogamous developments of the usual kind. With regard to their 

 nature it seems difficult to avoid the conclusion that they had originated from sixteen- 

 celled sporangia in which a minute proportion of viable but genetically unbalanced 

 spores had been formed. With a larger sowing the still smaller proportion of viable and 

 balanced types might have been detected. 



The importance of this experiment is that it shows at least one way in which saltations 

 can occur in spite of apogamy, although to what extent, if any, these have actually 

 occurred in nature is unknown. It also demonstrates the importance of the nucleus in 

 the determination of this type of apogamy, since here we have what is tantamount to 

 segregation of sexual versus apogamous characters in the oflTspring of one individual, and 

 the peculiar genetical behaviour of apogamy can therefore not be explained by any 

 type of maternal cytoplasmic inheritance. 



The genetical inference would appear to be this. Apogamy cannot be determined by 

 a simple Mendelian mechanism involving one mutant factor which is either dominant 

 or recessive. It seems rather to be the expression of a generalized unbalance of a quantita- 

 tive kind involving several (i.e. at least two and possibly many) different processes the 

 interaction of which is required for development of sporangia and prothalh, and some 

 or all of which are liable to fluctuate from environmental causes internal or external to 

 the plant. The exact fate of any individual sporangium is therefore to some extent 

 indeterminate, though the statistical frequency of the diflferent types of development may 

 be constant and characteristic for a given plant in a given environment. 



Such a condition could perhaps be produced in a pure line by the simultaneous 

 occurrence of two or more genetical mutations of appropriate type; such an occurrence 

 is, however, in a high degree improbable. For this reason it need cause no surprise that 

 no case has yet been recorded of apogamy of this kind occurring as a local variant 

 within a pure species. Multiple unbalance of many genetical factors can, however, 

 readily occur at one step by the mating of gametes of different constitution, and it can 

 be no accident that interspecific hybridization has been suspected or proved for every 

 example about which sufficient information is available. That such hybrids often in- 

 volve the mating of gametes of different grades of polyploidy is perhaps only incidental, 



194 



