1892. | Evolution in Methods of Pollination. 45 
ranged reproductive organs, such as the antheridial and arche- 
gonial clusters of mosses or the spore-bearing stalks of ferns 
and equisetums. The growth of bracts, i. e., leaves altered 
to do protective work, and the further development in the 
macrosporangium mouth of the mucilaginous secretion already 
foreshadowed in the archegonia of ferns, produced the char- 
acteristic inflorescences of gymnosperms. The position of a 
plant, as of a man, in the scale of progress, is measurable by 
the protection given to the children and by the manner of 
their preparation for independent life. The increase of ovule- 
shelter by the formation of a closed ovary is an easy step, as 
the comparison of the ripened pods of Mitella, Tiarella, Aqui- 
legia or almost any of the Leguminosz with the ovule-bear- 
ing scales of pines, shows. By the incurving of the edges of 
one of these scales, or more probably, by the persistence of 
the inrolling of the edges of the young leaf, an ovary perfect. 
in every essential would be produced, and the favorable varia- 
tion transmitted to succeeding gererations.! Still further 
Provision for the safety of the seeds and for their advancement 
in life, is attained by increased development of the protective 
bracts to form organs such as the perigynia of sedges, the 
rom the united tips of the carpellary leaves. 
Then the lords of horticulture, the insects, with an eye to 
Profit, began their investigations of the fields, at first obtain- 
'ng only pollen from these wind-tossed inflorescences. In 
Some cases they never find anything more, e. g., in Hepatica 
and Papaver. But the occurrence of sugary secretions, 
Pegg time after writing this sentence, I came across a remarkable age 
6 of the truth of the theory in Eichler's “‘ Bliithendiagramme,’’ part II, p. 
1». He says: ‘In most Resedaceae the carpels are so united that they form 
anoth Ovary with parietal placentae, . . yet they remain free pico 
th “r at the top, nor do the edges of the individual carpels close togethe' 
men teat the ovary is open above. . . . Th dition of things is 
oa different in Reseda luteola, Caylusea and Astrocarpus. In the first 
rect g Ae individual carpels remain separate, their edges turned inwards and 
low to bear the ovules in the ordinary way, but not touching above 
sang ; the ovary appears therefore to be made of five or six free scales w) : 
form.” ¢ .ircle-— a very primitive structure, not occurring elsewhere in t ~ 
The state of things in Tiarella cordifolia seems to me to correspon) 
well to this description of the pistils of Reseda luteola. 
