JfAUDIN — ON HTBBIDISM. 



5 



that of a large walnut, and some of them were very spiny, while 

 others were covered with tubercles, or almost destitute of spines ; 

 certain individuals bore fruit at the first fork, while others were 

 fertile only towards the last, and finally there were some which set 

 only a single fruit. On the whole, the forty-five plants of the two 

 sets constituted, so to speak, as many individual varieties, as if, the 

 bond which ought to unite them to the specific types being broken, 

 their vegetation had wandered in every direction. This is what I 

 call irregular variation, in opposition to another very different 

 mode of sporting, of which I shall speak presently. 



I could bring forward many other examples of the excessive 

 variability which arises in consequence of crossing. Not being 

 able to give to this note all the space which the subject demands, 

 I shall confine myself to the following examples, which have 

 likewise been furnished by my experiments. 



In 1863 I received from an horticultural amateur a full-grown 

 plant of Mirabilis longijloro-jalapa of the first generation, the 

 issue, as the name implies, of the common Marvel of Peru, with 

 purple flowers, fecundated by M. longiflora. A seed obtained 

 from the first cross of the two species accompanied the plant, 

 which was destined to give me a second hybrid individual, equally 

 of the first generation. The two plants, cultivated side by side, 

 became enormous. Intermediate in the same degree between the 

 parent species, which they far surpassed in stature, they resem- 

 bled each other as exactly as possible, which might be expected, as 

 they both belonged to the first generation. They were mode- 

 rately fertile, and out of the many thousand flowers which they 

 expanded in the course of nearly three months, they produced 

 some hundreds of perfectly similar seeds. 



The older of these two plants having already borne fruit the 

 preceding year, and some of its seeds having been sent to me by 

 the donor, I obtained the same year (1863) six other hybrid 

 plants of the second generation. None of them acquired the 

 large stature of the hybrids of the first generation ; none, more- 

 over, resembled them. Of these six plants, there were two which 

 seemed to be the image of each other, so slightly did they differ : 

 this was an exception; they flowered abundantly, but though 

 well developed and very vigorous, they remained entirely sterile. 

 A third almost reverted to the normal form of Mirabilis jalapa, of 

 which it possessed the stature, the leaves, the flowers, and the fer- 

 tility ; it differed only in a slightly more expanded habit, and the 

 tube of the corolla being more elongated. The three last were 



