82 BOTANICAL GAZETTE [january 



Rettig, Ule, 10 and others, all of which are out of harmony with the myrmecophile 



hypothesis. The work of Nieuwenhuis-Uexkull confirms these more recent 

 views. 



After a detailed account of extra-floral nectaries by plant families, the author 

 summarizes the data presented, and some of the chief conclusions follow. The 

 structure and form of the nectaries do not favor the theory that they originated 

 as adaptations for ant protection; in many cases they specifically oppose such an 

 assumption, and their position on the plant (largely on the leaf undersurface) 

 is such as to be of no purposive significance. The secretions often begin late in 

 life, so that the plant is without protection in youth, when it is most needed. In 

 other cases the secretion begins in early youth and soon ceases, thus leaving the 

 plant for a long time without ant protection, if such exists. The nectaries usually 

 secrete sugar somewhat spasmodically during their period of activity, and are often 

 . dry. The nectar of many species is avoided by ants and other animals. The 

 view that the honey-seeking ants drive off crawling insects and other "unbidden 

 guests " that mutilate the flowers, robbing them of honey or pollen, is quite 

 untenable, there being no relation between mutilated flowers, ants, and extra- 

 floral nectaries. Floral mutilation depends on the structure and position of the 

 . flower or the weather; furthermore, most mutilated flowers produce as many seeds 

 as flowers that are not mutilated. The honey-seeking ants are not combative 

 and do not attack other insects on the plants they visit; indeed, these other insects 

 often attack and repel the ants. The nectaries, therefore, so far from being bene- 

 ficial structures developed by natural selection, are harmful to the plants of which 

 they are a part, in that they attract insects of all kinds, which not only eat the 

 sugar but do harm in various ways. Observation showed that individual plants 

 which secreted little or no nectar are less harmed by insects than are those that 

 produce nectar. 



flement 



rmeco 



v _ 7 v ivw.au. UiiVUl^ I^V-» U1VUUVU aa-v^^*» ~ 



literature. Ants may "love" plants, but there is no evidence that plants "love" 

 ants. Plants inhabited by these insects, if it seems worth while to group them, 

 may be called myrmecophytes — H. C. Cowles. 



A Mendelian ratio and latency.— Shull" in a suggestive paper makes 

 further contributions to Mendelian theory. In certain bean hybrids thre'e distinct 

 units were shown in earlier papers 12 to be involved, namely, a pigment factor, a 

 blackener, and a mottled pattern. In the last character a peculiar condition is 



depe 

 10 See Box. Gazette 44:314. 1907. 



allelo 



Shull 



Amer. 



908 



12 



, The significance of latent characters. Science 25:792. 1907; S ° mc 



latent characters of a white bean. Idem 25 1828. 1907. 



