382 
AMERICAN JOURNAL OF BOTANY [Vol. 9, 
tations to insure cross-pollination. Although the inflorescences of certain 
Marcgraviaceae are known to be visited at times by insects and birds, 
there is no reliable evidence to indicate that these animals actually are 
concerned in the pollination of the flowers. Statements to the effect that 
certain species are entomophilous or ornithophilous are based upon purely 
teleological inferences, rather than upon detailed field observations and 
carefully planned experimental controls. Furthermore, the prevailing 
view that the Marcgraviaceae are protandrous may be traced back to 
Delpino (1869), whose conclusions appear to have been derived, not from 
the study of living plants, but from the examination of herbarium material 
and from Martins' figures of certain species of Ruyschia and Norantea. 
It is essential, therefore, that various species of Norantea, Ruyschia, 
Souroubea, and Marcgravia be critically studied in order to determine 
(i) whether any of the Marcgraviaceae are protandrous, and (2) whether 
the insects and birds which visit the nectaries actually are concerned in 
cross-pollinating the flowers. 
Delpino (1874), Belt (1874), Kerner (1876), and certain of their con- 
temporaries were of the opinion that extra-floral nectaries are adaptations 
for attracting insects or other animals. The theories of these investigators 
have been severely criticized, if not actually demolished, by Rettig (1904), 
von Uxkiill-Giildenbrandt (1907), von Ihering (1907), and others, and it 
must be admitted that biologists are still as ignorant as they were in the 
days of Linne concerning the true function of the extra-floral nectaries 
and the so-called food-bodies of plants. Therefore, it is to be emphasized, 
in conclusion, that the hypophyllous glands and nectariferous appendages 
of the Marcgraviaceae, and extra-floral nectaries and ''food-bodies" in 
general, deserve to be studied intensively along physiological lines. 
Conclusions 
1. Although the inflorescences of Marcgraviaceae are visited at times 
by insects and birds, there is no reliable evidence to indicate that these 
animals actually are concerned in the pollination of the flowers. 
2. The highly specialized inflorescences of Marcgravia umhellata L., M. 
ciiyuniensis spec, nov., M. purpurea spec, nov., and of similar species do 
not appear to be efficient mechanisms for insuring cross-pollination by 
humming birds. The pedicels and nectaries are so arranged that birds 
tend to approach the inflorescences from above and, therefore, do not 
become coated with pollen which subsequently is rubbed off on the pistils 
of other flowers. 
3. The flowers of the only two species of Marcgravia, M. (Myuniensis 
and M. purpurea, which have been studied in detail in the field, appear 
to be self-fertile or autogamous, instead of being protandrous and cross- 
pollinated by birds. 
