1916] DEVRIES—DIMORPHIC MUTANTS 270 
were cana and 23 were Lamarckiana, while one dwarf occurred. 
The percentage figures, 52 cana and 46 Lamarckiana, confirm those 
of table X. 
The reciprocal cross yielded only normal Lamarckiana, 60 
specimens with one dwarf. Of these 23 have flowered. 
O. lataXO. cana.—In this case the differences were already 
very clear at the beginning of June. Therewere 34 lata, 61 Lamarck- 
tana, and 3 mutants (one each of cana, nanella, and oblonga). In 
August 7 lata and 9 Lamarckiana flowered. They repeated the . 
type of the hybrids of O. Jata with the parent species. 
O. canaXO. nanella—Only 24 seeds germinated. These pro- 
duced 5 cana, 17 Lamarckiana, and 2 dwarfs which had the char- 
acteristics of O. cana. All these plants have flowered. 
The reciprocal cross gave 111 plants, of which one was a cana’ 
of normal stature, 105 were cana with the dwarfish stature of 
O. nanella, and 5 had the type of O. Lamarckiana. Two of this 
last type, the normal specimen of cana, and 22 dwarfs flowered in 
August. 
O. oblongaXO. cana.—Already in June the differences were 
clear and unmistakable. Among 72 plants 59 were of the Lamarck- 
tana type, 11 oblonga, 1 albida, and 1 rubrinervis. Of these, 2 
oblonga and 12 Lamarckiana flowered; most of the others lived 
through the summer in the condition of large rosettes of radical 
leaves. , 
O. canaX O. rubrinervis —Only 20 seeds germinated, and yielded 
8 cana of the normal type, 5 Lamarckiana, and 7 subrobusta, all of 
which have flowered. 
The reciprocal cross yielded 59 plants, but in n only two types, 
which were the same as those in the first instance. Of these, 17 
Lamarckiana and 8 subrobusta have flowered. 
Summary 
1. Besides O. scintillans, which splits under ordinary circum- 
stances in every generation into nearly equal groups of plants of 
the same type and others of the type of O. Lamarckiana, I have 
cultivated pedigree families of 4 other mutants of O. Lamarckiana 
which behave in the same manner. They have been designated 
