152 Report S.A.A. Aijvanckmknt of .Science. 



in all diiections, and lose the good qualities for which we have 

 specially paid. 



Dr. N. Hjalmar Nilssoii, of Swalof in Sweden, seems to have 

 achieved remarkable success by establishiiii;- pure races of cereals 

 through breeding from single inflorescences. In this way he seems 

 to have solved the difficulty of having to guard against deterioration. 

 But how can we liarmonise these facts with the theory of the relative 

 constancy of species ? However )nodifled a barley may be through 

 Dr. Nilsson's selections, I doubt whether we are justified in looking 

 upon it as specifically different from Hordewm KO.tifum, L., or more 

 properly the wild Hordeuiii ><n<yntaneum, C. Koch. I fully expect 

 that all the races of barley will, when tested by the l)io]ogical method, 

 give the same reaction, just, as far as we know, every member of 

 the species Homo sapiens gives the same reaction. If we let our 

 imagination have some play again, I think we shall have to look 

 upon each species as a system which we may louglily compare, to a 

 certain extent, to the sun with its planetary sytem. The sun. repre- 

 senting the essential specific nucleus, has immense stability, while its 

 satellites, representing minor characters, ma}' change : some maj' be 

 lost, while others ma}' be added to it, and thus the aspect of .the 

 whole system may be altered without losing its essential character- 

 istics. I am (|uite prepared to defend more in <:letail this view, which, 

 in plain words, makes a distinction between truly specific and varietal 

 characters of species — a distinction, moreover, already admitted by 

 De Vries. He even asserts that the latter alone exhibit Mendelian 

 characters. We require only to go a tiny step further to enable us to 

 make much of his tlieory tumble to pieces, at least as far as it is based 

 on the supposed formation of new species, as it were^ under his (;wn 

 eyes, while leaving his fundamental conception intact. 



However, this whole subject is so wide, that I may well be pardoned 

 if I leave it in this sketchy state. No doubt the next few years Avill 

 bring us many new' facts which will advance it both from a purely 

 theoretical and a practical point of view. Plant breeding is a com- 

 paratively new science, but it is now practised in mnnerous purely 

 scientific establishments as well as by many practical men. Amongst 

 the latter Mr. Luther Burbank stands easily first. This marvellous 

 man has received full recognition froni scientific men like Hugo 

 de Vries, though he has sufiered much at the hands of his friends, 

 who have "written him up" in true American newspaper style. 

 Hybrids form to him the starting-point for selection. He uses 

 enormous numbers. Hundreds of thousands of specimens are some- 

 times raised to yield a few selected types, or to be discarded altogether. 

 By dint of practice Burbank is said to be able to tell from seedlings of 

 certain fruit trees the (juality of the fruit they will bear. If this is so, 

 and I have no reason to doubt it, then we have here eori-elations which 

 perhaps one day will yield laws that are at present more or less hid- 



