vii Relations Between Plants and Animals 133 



this are numerous pear-shaped or oval little bodies which 

 look like insects' eggs. These (" Miiller's bodies ") are 

 eaten by the ants, and they must be very nutritious, for their 

 contents are rich in albuminoids and fatty oil. In their 

 young stages they resemble the glands of other plants which 

 secrete mucus or resin, and Francis Darwin has suggested 

 that they are homologous with glands, but they are very 

 peculiar, and their peculiarities Schimper would connect 

 with the ants, for, strange to say. they are absent on 

 Cecropias, such as those of Corcovado, on which there are 

 no ants. 



From Cecropia Schimper passes to other myrmecophilous 

 plants, such as Acacia sphtzrocephala, whose hollow thorns 

 afford shelter to a bodyguard, and which also bears little 

 food -bodies comparable to those on Cecropia. But we 

 have given sufficient illustration of this matter. 



There is, however, another much-discussed subject on 

 which Schimper has something interesting to tell us. We 

 know that many of our flowers have honey-bags or nectaries, 

 and that these are attractive lures to the bees and other 

 insects which, in their search for sweets, carry the fertilising 

 pollen from flower to flower. But often, especially in the 

 Tropics, nectaries occur outside the flowers, as we have 

 seen in Nepenthes. 



Both Belt and Delpino regard these extra-floral nectaries 

 as adaptations for the attraction of protective ants. Kerner 

 supposes that they afford such generous supplies of nectaries 

 that the ants leave the floral nectaries undisturbed ; but this 

 is not the way with ants! Bonnier regards them as stores 

 of reserve-material ; but the sugar is almost always stolen by 

 ants or washed away by the rain. Johow has even suggested 

 that they are receptacles for the waste-products formed in 

 the movements of the leaves! but this too is for several 

 reasons most unlikely, since it is enough to notice that 



