proper detail. Dr. Dall has already traced more or less completely the genealogy of some of the species, and I have noticed certain series of species—the group of Corbula fossata, C. oniscus, C. wailesiana, etc., being one of them—deserving thorough study, but the paleontologic work known to me has not as yet been done with the requisite detail to form the basis of an opinion. The principal contribution to our general knowledge of the evolution of organisms that paleontology can make, however, is, I believe, in tracing out phylogenetic lines, and I believe the discovery of the processes and causes of evolution must rest with the experimental biologist. During the past few years very important experimental investigations have been made by several men, and I venture to refer to their results, as I regard paleontology as only an aspect of biology, and think the students in that field should utilize the information gleaned in others.
In the study of variation it has been shown that the selection of fluctuating variations does not carry the species beyond a certain limit, or the extent of the variation is limited, leading to the conclusion that new species can not be produced by this method. I may here refer to ecological surveys and the unreliability of conclusions reached by such researches. Dr. Merriam several years ago presented a paper "Is Mutation a Factor in the Evolution of the Higher Vertebrates?" in which he announced the conclusion that it was not. A critical examination of Dr. Merriam's data showed he had not sufficient information on which to base such a conclusion. His data possess value for the study of evolution in that they indicate material that may be profitably investigated by the experimental method. Attention should also be called to the probable insufficiency of conclusions reached by studying material from successive geologic horizons. For instance, suppose that two usually distinct forms are connected by intermediates. There are no means of ascertaining whether the intermediates represent transition stages between the two forms or are examples of blended hybridism.
That new species may originate through saltation is rather definitely proved; but that it is the only process is not established.
To consider the causes of the origin of new forms: That new forms should originate from the old without the action of some new influence seems to me impossible. The circle of possible combinations of already existent characters could not be transcended, and there would result by crossing only all the combinations possible within definite limits; this would be especially obvious if the de Vries hypothesis of unit-characters be true. Many experiments to determine the influence of various physical factors on individuals showed only somatic changes not of heritable nature and the data accumulated seem definitely to prove that somatic changes, or acquired characters induced through the soma, are not inherited.
