194 PRINCIPLES OF PLANT-TEEATOLOGT. 



hybridum) the carpel of some flowers had become en- 

 larged, green, and more or less open, otherwise hardly- 

 changed. In the Dutch clover {T. repens) the carpels 

 in some flowers were foliaceous, but showed differing 

 degrees of development, for in some the leaf so- 

 formed was unifoliolate, in others bi- or trifoliolate 

 (PI. XL VIII, figs. 4 and 5). This virescence of the 

 carpels in this species is sometimes accompanied by 

 that of the sepals and petals and atrophy of the 

 stamens. Molliard has shown that virescence in this 

 plant is due to the presence in the stem of the larva 

 of an insect, probably of Hylastinus obscurus, which 

 forms galleries in the pith and sometimes penetrates 

 to the cortex. In this way the proper circulation 

 of the sap was prevented. Every abnormal plant 

 examined was shown to contain a larva, while none 

 could be found in healthy plants. 



In the melilot {Melilotus arvensis) the virescent 

 flowers were found to be due to the presence of the 

 larva of (probably) Apion melilotii living in the neck 

 and branches of the stem. No larvae could be found 

 in normal plants. 



In the rose the carpels may become foliaceous while 

 not as a rule reaching the development of the foliage- 

 leaves, remaining simple instead of compound, but 

 this nevertheless constitutes a great change from the 

 normal achene; such leafy carpels also become su- 

 perior in position. In the " green rose " (Bosa indica 

 var.) and in proliferated and fasciated roses this 

 change can be seen ; in proliferated roses the modified 

 carpels often form the lower leaves of the shoot which 

 higher up exhibits transitional forms between them 

 and ordinary foliage-leaves. 



In the double flowers of the cherry, where two 

 carpels are present, these often exhibit interesting 

 intermediate stages between carpels and green leaves, 

 the midrib being prolonged into a short style, termi- 

 nated by an imperfect stigma. In very many normal 

 pistils the original marginal placentation has become 



