8o PROCEEDINGS OF THE 



experimeDtal ba,sis of the mutation-theoiy is lost, for de Tries 

 expressly states : — 



" I have in the course of the years cultivated- many more than 

 a hundred of such species, but essentially only one has iu fact 

 come up to my wishes." 



This one was (E. Lamarckiann. 



I also know of no other case in which progressive mutation 

 from a pui'e species has been shown to occur, so that I feel 

 iustified to say that so far there is no proof of the existence of 

 progressive mutation. 



Now about the question of 



Mutation by Loss of Factors. 



The existence of such mutations would not, as pointed out, be 

 fatal to my theory. I even acknowledge the possibility of such, 

 a loss occurring accidentally, though I do not tliiuk it has been 

 proved. As the question is of no vital importance, and as I have 

 so little time, I will say very little about it. 



At first sight, there seems to be plenty of evidence for this, but 

 it nearly always is of a circumstantial kind. In this respect a 

 role is played by the old belief in which we have been brought up, 

 that of a certain number of forms resembling one another, one 

 must be the original or at least be nearest to the original one 

 from which the others arose. This leads to the involuntary 

 conclusion that the most common form is the original and that 

 tlie rarer ones are derived from it ; so that, if we find a glabrous 

 form among a large number of hairy ones, it is at once assumed 

 that the glabrous one arose from the hairy one by loss of a factor. 

 This, however, is a mere assumption. The aberrant form may 

 just as well owe its peculiarity to a former cross. 



A very nice case proves that this standpoint is well founded. 

 Brainerd, the well-known investigator of American violets, 

 describes a case in point. Viola palmata is a pubescent plant. 

 Near Spring A-^alley, in the State of New York, a glabrous 

 specimen referred to Viola ixdmata as a variety was found. Self- 

 fertilized seeds of this specimen gathered after its transplantation 

 in Brainerd's garden were sown, with the result tliat segregation 

 was observed in plants w-ith the deeply lobed leaves of V. pahnaia 

 and such with the uncut leaves of V. jyapilionaiea. The apparent 

 V. palm ata, \2ir. glabra^ was thus shown to owe its origin to a 

 cross with V. papilio'nacea. Later on, Brainerd found at Tvron, 

 North Carolina, very far from the first find in New York, 

 therefore, another colony of a glabrous Viola palmata, which in 

 the same way was proven to be a hybrid with the glabrous 

 V. papilionacea . 



This case alone, 1 think, shows that for the proof of tlie 

 existence of mutation by loss of factors the purity of type from 

 which we start is the very first requiiement. 



Hagedoorn thinks lie knows of the existence of seven cases ot 



