72 CARNEGIE INSTITUTION OF WASHINGTON. 



Perennation of the above-ground shoots is here possible. At Tucson, 

 a lax, even prostrate, habit has become evident, the leaves appear 

 nonnal in color and shape, and flowering and fruiting are abundant. 

 In both situations there is considerable thickening of the leaf, expressing 

 itself both in an increase in the number of layers of cells and in their 

 enlargement and subdivision. Since these characters are appreciable 

 and measurable, it becomes possible to test the effect of environment 

 on the soma in terms of inheritance. 



Floral behaviors. — At Carmel the climatic conditions of the late 

 spring and summer are unfavorable for the development of flowers. 

 Flower-buds are formed, and the inflorescence may appear, only to be 

 arrested before any anthesis. Inflorescences 2 cm. long have been 

 observed to remain unchanged for over three months, while the shoots 

 bearing them advance in growth. During the present summer only 

 one inflorescence normal in form and color was produced, but in this 

 the perianths did not spread open, though the stigmas were exposed and 

 pollinated. In spite of the latter circumstances, there is no indication 

 that seeds will be produced. Plants growing under a well-ventilated 

 glass shelter produced normal flowers, and the prompt enlargement of 

 the ovaries indicates that seeds will develop. This shelter preserves 

 the temperature a few degrees centigrade (2 to 4) above the usual ones 

 prevalent on foggy days, and gives protection from the cool winds. 

 It may be concluded that the failure to form normal flowers and seeds 

 during this time of the year results from the too lov/ temperatures. 



Marked departures from the normal have been shown by two out of 

 five old plants which have been allowed to persist at Tucson. One of 

 these (No. 2) in 1913 produced only abnormal inflorescences, in which 

 the flowers remained in the form of minute vegetative buds in the axils 

 of the bracts and bracteoles, which were much hypertrophied, and the 

 branching system had been advanced from simple to compound. 

 Another plant (No. 1) showed similar structures on a single basal 

 branch in 1913. 



The shoots formed in 1914 on these plants showed additional alter- 

 ations, consisting chiefly of insect galls; the double effect was exhib- 

 ited by every inflorescence on plant No. 2, while a few normal racemes 

 were formed on No. 1. The gall is due to the stimulation of the 

 pistil by an unidentified and possibly new insect. Indeed, two insects 

 have been found within the galls, one cynipidous, the other gnat-like, 

 in separate locules of the same pistil. There is also always present 

 a fungus which lines the cavity, and the frequent occurrence of mum- 

 mies of larvie indicates that the fungus is parasitic on the insect or 

 insects and not on the plant, since it is not found in otherwise un- 

 occupied ovaries. The inference that this whole abnormal behavior 

 is referable to insect stimulation is possible, but is not supported by 

 the evidence at present available. For example, very young inflores- 

 cences in which total suppression of flowers has occurred show no trace 



