488 On the Latent Capacity for Mutation. 



(1887). Cases are not rare in which supernumerary 

 petals are formed by the transformation of one longi- 

 tudinal half of a stamen into that of a petal. I found 

 cases of this at Hilversum in 1886 and 1887, and after- 

 wards in my cultures. Usually only one such organ is 

 present in a flower but sometimes more, and once I ob- 

 served as many as four (1894). There also occurred 

 stamens which took on the form of petals by both the 

 filament and the anther becoming flattened (1887, 1888 

 and later in the cultures). Cases of the fusion of two 



-5iSi 



Fig. 112. Oenothera Lamarck- 

 iana. Buds in the forks of split 

 cotyledons. 



Fig. 113. Oenothera Lamarck- 

 iana. Bud in the fork of a split 

 cotyledon. The bud has grown 

 up to a strong lateral rosette; 

 its base has much swollen in 

 consequence. 



filaments together, and of the fusion of a filament with 

 the pistil are fairly rare. I observed the former in 1887, 

 the latter in 1894. 



As the foregoing summary shows, the common floral 

 malformations occur in O. Lamar ckiana. I mention 

 them briefly because I have not laid great stress on ob- 



