230 THE NATURAL HISTORX BEYIEW. 



mention of cross-impregnation, or of insect action in plant fertiliza- 

 tion. It would appear that the possibility of either has never for 

 a moment been present to his mind. With the recognition of the 

 powerful influence of insect agency, the whole edifice, raised with so 

 much paius, tumbles to the ground. The facts may stand, but they 

 acquire a different meaning, and lead to quite different conclusions. 

 Every one knows how little reliance can be placed on the results of 

 experiments, in the course of which a great number of closely allied 

 forms are grown together. We are all familiar with the effects of 

 the " visits of bees," and know how difficult it is, even with a net, 

 *' to keep out small diptera." 



The conclusions drawn from observations on seedling plants 

 grown promiscuously must therefore be set aside altogether, as at the 

 best unsatisfactory and unconvincing. We are quite Availing to be- 

 lieve, on M. Jordan's authority, that the different closely allied 

 forms re-appeared year after year. How and why this was so is a 

 curious matter for enquiry, but there is nothing at all to show that 

 it was because each form was fertilised by the pollen of the same 

 kind and no other. When a peculiar variety is cultivated alone, 

 with no nearly allied race or species near it, general experience 

 confirms the result obtained by Jordan, of the retention of the 

 characters during a certain number of generations, as many as 

 twenty-five in the experiments before us, a number considerably less 

 than infinite. The third result, the permanency during a short 

 series of years of forms removed from their native place to a garden 

 is also conformable to observation. We are aware that it was at 

 one time thought that such variations were accidental and not 

 hereditary, and that change of locality by removing the cause would 

 effect a return to the normal state of the species. The permanence 

 of slight variations under cultivation has therefore been often 

 appealed to as a proof of specific difference. No doubt any modifi- 

 cation of character produced by external causes would disappear 

 with the removal of the cause, just as the changes produced by 

 excess of nutriment, aff'ecting only the luxuriance of the plant, con- 

 tinue only as long as the rich food is supplied. Even these slight 

 varieties are, however, now commonly believed to have a tendency 

 to become hereditary, without being therefore necessarily of specific 

 value. 



On the whole, a careful study of M. Jordan's work leads us to con- 

 clude that its greatest fault is a want of precise details of the experi- 



