SYMBIOSIS 143 



pollinators to pierce the corolla tube and secure the nectar without 

 rendering the proper return, from which a chain of coactions may be 

 set up. In California, the tanager punctures the thick tube of Big- 

 nonia cherere, and various species of hummingbirds that can neither 

 do this nor obtain the nectar directly profit by his fondness for 

 sweets, as do the ants and other insects that follow. On the other 

 hand, nectar may rarely be poisonous, as in Aesculus californica, and 

 tlie symbiotic relation leads to tragedy. 



These direct effects do not comprise the whole relation between 

 flowers and insects, or other pollinators such as the birds. The experi- 

 mental study of pollination has brought much objective evidence to 

 the support of the views of Darwin, IVIueller, and others, who believed 

 that the two partners have mutually affected the evolution of each 

 other. Flower form and color are most intimately linked with the 

 structure of the preferred visitor, while the form and size of the 

 latter have been reciprocally molded by its flower preferences or by 

 its collecting behavior, as exemplified by the so-called pollen baskets 

 of bees (cf. Clements and Long, 1923; Clements and Clements, 1928). 



The community consequences of this type of mutualism may be 

 inferred from the usually greater size and vigor of crossed plants and 

 their enhanced seed production. These not only insure a steady and 

 even an augmented supply of nectar and pollen for the animals con- 

 cerned, but also a tangible increase in the food supply of all gramin- 

 ivores and frugivores, from weevils to birds and rodents. This is 

 reflected in the larger number of seeds that escape destructive food 

 coactions and hence are available for germination and the maintenance 

 of the species. 



Animal Symbionts. The question as to the presence of symbiosis 

 in many animal coactions has been complicated by observation of a 

 general nature and the not infrequent injection of prepossession (cf. 

 Step, 1913). It is manifest that a very large number of assumed 

 symbioses, and especially those that merely involve attachment, shel- 

 ter, or lodging, are not at all mutualistic or only to such a slight 

 degree as to be insignificant. Even in the fresh-water mussels 

 (Anodonta, Unio, etc.) and the bitterling (Rhodeus), the relation ap- 

 pears to be a matter of reciprocal parasitism of a sort rather than 

 symbiosis, at least in the sense indicated by Step {loc cit., page 83). 

 This interpretation is supported by the fact that there is no such 

 mutual relation with other fishes that serve as hosts to the glochidia. 



]\Iany examples of symbiosis have been drawn from the social 

 insects, especially ants, but only a relatively small number of these 

 are mutual in any important degree. Most of these are covered by 



