72 Universih/ of CaJifnrii'xi riihlicdlimis in Bolanjj | Vol. 7 



called to an ohscrvatioii by Peek (1885) who. in describing the usual 

 long-petioled, green-striped variation, says that "the specimens indi- 

 cate a coincidence between the petioles, peduncles and green color of 

 petals. Generally the longer petioles are accompanied by longer 

 peduncles and broader green stripes on the petals. The coincidence 

 between form and structure is tiuly remarkable." 



Holzinger (1901) figures a plant in which the leaves are normal 

 but the flower consists- of six whorls of three green, foliaceous perianth 

 segments each. Osband (1894) describes an unusual plant, the flower 

 of which exhibited two whorls of three normal sepals each ; similarly 

 two whorls of petals, the outermost being striped or w'hite; and three 

 stamens, of which only one was normal, one being "expanded into a 

 half-sized petal" and anther-bearing, and the third sterile. The 

 ovary is reported as normal in size but "near the top of one of the 

 carpels arises an outgrowth about half an inch long, white, doubled 

 together and drawn over at the top like a hood." On the edges of 

 this outgrowth there were "two pollen-bearing lines about an eighth 

 of an inch long. ' ' 



Pollard (1901) reprints the figure and description of a plant noted 

 by Mrs. Kellerman (1893). The plant bore a conspicuous double 

 flower, and the double condition reappeared for some years from the 

 original rootstock, w'hich was brought into a garden. The flowers 

 were white in color and varied in number of whorls of petals from 

 nine to thirteen. The reproductive tissues were almost completely 

 suppressed. Smith (1879) comments upon the number of abnormal 

 individuals and their wade distribution in ^lichigan. Nothing is added 

 to the types noted by Britcher, but the statement is made that in 

 abnormal plants "the ovules are sometimes twelve in number, but 

 generally less, often none, or only one or two, the cavity being filled 

 by an enlarged placenta." It might be mentioned in this connection 

 that rootstocks of normally fruiting plants of TiilJium sessile var. 

 giganteum H. & A. when brought under greenhouse conditions have 

 apparently no dormant period (cf. Brandt. 1916), and each of the 

 flowering shoots matures a large ovary which contains only a spongy 

 mass of placental tissue. With regard to the prevalence of abnormal 

 individuals in ^lichigan, J\leehan (18946) reports that many ])laiits are 

 found with "four leaves, three sepals, and four petals, a less number 

 with four leaves, three sepals, and five petals. ' ' Since these forms are 

 found only at the end of the season, a causal relation is suggested 

 between lateness of flowering and abnormal variation. 



