REVERSION IN COMPOSITES 



The Sudden Appearance of Far-distant Ancestral Types of Inflorescence 



J. L. Collins 

 University of California, Berkeley, Calif. 



FOR several years the writer has 

 studied the morphological varia- 

 tions in Crepis capillaris, a cosmo- 

 politan plant belonging to the chicory- 

 group of the Sunflower family. During 

 this time seeds have been secured from 

 many localities of Europe, Asia, Africa 

 and the American continents. In seeds 

 from such diverse sources one would 

 expect to find considerable morphologi- 

 cal variation and such has proven to be 

 the case. 



In an Fi population secured by 

 crossing a plant of a Holland strain, 

 which had grown in our garden at 

 Berkeley for three years, with a plant 

 of a Sweden strain which had grown in 

 our garden for two years, there ap- 

 peared a single plant having foliaceous 

 palea-like bracts subtending the 

 achenes on the receptacle of every 

 capitulum. The normal condition in 

 these plants shows a perfectly smooth 

 and naked receptacle. The appearance 

 of these palea-like organs is considered 

 as a possible reversion to a pre-com- 

 posite type of inflorescence because the 

 phylogeny of the composite form of 

 flower aggregation indicates that the 

 progenitors of the family possessed 

 such structures as a part of their 

 inflorescence. There are at least tv/o 

 possible lines of development of the 

 composite capitulum ; one coming from 

 a spike-like inflorescence and the other 

 from an umbel or a racemose umbellate 

 type. In a typical spike the individual 

 flowers or spikelets are usually well 

 spaced apart, rather large and each 

 subtended by a bract, a condition simi- 

 lar to that which we now find in the 

 grasses, cereals, etc. Reduction in the 

 length of the central stalk of the in- 

 florescence proper (in a spike termed 

 the rachis) would cause the inflores- 

 cence to become more compact and at 

 the same time cause loss of some of the 



subtending bracts as a result of the 

 crowding; the end result would be an 

 approach to the composite capitulum. 



However, according to the opinion 

 of James Small, ^ the receptacle from 

 such a source, would rarely if ever 

 become entirely flat but would tend 

 to assume a rounded, conical shape 

 such as the fleshy receptacle of the 

 strawberry or the compact inflorescence 

 of Dipsacus, the common teasel. 



On the other hand, the flowers in an 

 umbel are already crowded and thus 

 reduced in size, the outer florets receiv- 

 ing more illumination and space caus- 

 ing them to become zygomorphic, a 

 condition found in almost all compos- 

 ites exclusive of the Cichoraceae. The 

 pedicels at their insertion on the 

 peduncle are more crowded than the 

 florets at the top of the umbel so that 

 the bracts subtending the pedicels 

 of all the flowers except the outer whorl 

 have already become much reduced or 

 have disappeared. Now if complete 

 abortion of the pedicels occurred the 

 florets would then become seated 

 directly upon a more or less flat recep- 

 tacle with much reduced inner bracts 

 or none at all. The parts of the calyx 

 of the outer whorl of florets would then 

 form the involucre while a shortening 

 of the peduncle, or stalk, below the 

 inflorescence would cause a transfor- 

 mation of cauline leaves into the 

 calyculus, subtending and partly en- 

 closing the involucre, which is a more 

 or less general character throughout 

 this family of plants. 



A study of proliferation of the in- 

 florescence such as one may occasion- 

 ally find in some genera of the family 

 (Crepis and Hypochaeris) supports the 

 umbellate origin hypothesis (Fig. 18). 

 In these cases, instead of the sessile 

 florets being produced on the receptacle, 

 pedicels are formed, each of which will 



Small, James. Origin and Development of the Compositae. London. 1919. 



129 



