8S tEGETABLK HYBRIDITY. 
F 
I 
loration of the stem and flowerSj in fertility, and in tlie size and supe- 
riority of the fruits. One plant of the Icevi-ferox series had completely 
reverted to the type of D, lavis, except that the base of its stem still 
bore a violet ring ; a few showed faint traces of resemblance to D, 
ferox, but the majority were more like Z). Stramonium and querclfoUa^ 
-with which they had no relationship, than the species from which they 
were descended. "In fine," says the author, after describing some of 
their chief differences, " the forty-five plants of the two lots formed, 
so to speak, as many individual varieties as if, the connection which 
shoidd have bound them to specific types having been broken, their 
vegetation had deviated in all directions. This may be called disor^ 
dered variation in opposition to another and very different mode of 
var}ang which will be mentioned hereafter." 
In 1863, M. Naudin obtained a plant and seed oiMirahilis longifora- 
Jalaipa of the first generation, procured by impregnating the common 
pnrple-flowered Man^l of Peru with pollen of M. longijlora. The seed 
was sown, and the two plants grew to a large size, perfectly similar in 
every respect and intermediate between the parent species. They were 
moderately fertile, and furnished some hundreds of perfect seeds. 
From seeds of the first plants obtained in 1862, M. Naudin raised 
six other hybrids, of course of the second generation. These did not 
resemble the hybrids of the first generation either in size or appear* 
ance. Two of them were nearly alike ; they were vigorous, and 
flowered abundantly, but were quite barren. A third had almost re- 
verted to the if. Jalapa, differing chiefly in the longer tube of the co- 
rolla; this was fertile. The remaining three were stunted in their 
growth, very dissimilar in appearance, and barren, — at least they pro- 
duced only a few fruits in which the seeds were imperfectly formed. 
Three new plants of the second generation grown in 1864 presented 
the same diversities; they resembled neither those of the preceding 
year nor the first hybrids. One of them, which approached M. Jala^a 
in its characters, was very fertile; the others flowered irregularly and 
were barren. This second experiment gives further evidence of the 
disordered variation of the products of a liybrid plant, when they do 
not revert towards one of the parent species* 
It becomes a question whether this tendency of the hybrids to vary 
continues to the third and following generations. In 1863 and 1864 
the author observed the sixth and seventh generations of a hybrid,. 
