74 PAEASITES BELONGING TO THE GENUS GLOMEEELLA. 



other parts of the colonies (PL XV) ; but in no case have perithecia 

 been produced under such conditions when the conidia belonged to a 

 race of the fungus which did not produce perithecia in the separate 

 colonies. Edgerton's (32) experiments along this line already- 

 referred to are not conclusive. That no such union of different 

 strains is necessary or essential for the production of perithecia seems 

 certain from the fact that perithecia develop in abundance in a colony 

 derived from a single conidium or ascospore of a perithecial race. 

 Until it has been demonstrated that the increase of perithecia at the 

 point of contact between two colonies is due to nuclear fusions 

 between the two growths it seems preferable to attribute the pheno- 

 menon to some simpler cause such as the greater exhaustion of the 

 nutriment at the point or some other slight stimulus. For the present 

 we must admit our utter ignorance of the determining factor or 

 factors concerned in the production of ascogenous fructifications in 

 Glomerella. Certain previously supposed factors, however, such as 

 have already been discussed here, have in most cases been suffi- 

 ciently tested to justify their elimination. This helps to simplify 

 the problem somewhat and suggests research in other directions 

 which may possibly prove more profitable. 



INOCULATION EXPERIMENTS. 



Inoculation experiments with Glomerella, either in the field or 

 greenhouse, are not always conclusive. Under proper conditions and 

 with plants known to be free from dormant infections, inoculations, 

 by surface application of spores, under optimum conditions of tem- 

 perature and moisture, should give the most trustworthy results. 

 To be certain, however, that a plant is free from dormant infection, it 

 must be grown from seed under conditions which would preclude the 

 possibility of infection from any source except the inoculation. 



Most of the inoculation experiments of the writers have, however, 

 been made with fruits or plants growing, under ordinary laboratory or 

 greenhouse conditions. As it is a practical impossibility to determine 

 whether any particular plant or part of a plant is entirely free from 

 infection, under such conditions, the results obtained can not always 

 be regarded as conclusive. In most of the experiments described here, 

 inoculations were made by insertion of conidia or ascospores of Glome- 

 rella in the tissue of the host. Such experiments with immature 

 fruits when made under proper conditions and with sufficient checks 

 are believed to give evidence of some value in determining the host 

 relationships of the organisms. In the case of perfectly mature fruits 

 the significance may not be so great. 



Since it has been found that different races or strains of Glomerella 

 vary exceedingly in virility and also that different host plants and 



252 



