obstacles that we do not attain the ideal. The practical question is, What are these obstacles, and how may they be removed, avoided, or overcome? We undoubtedly make many failures, which are inevitable, and for which we are not responsible; I mean such failures as are due to the present limits of scientific knowledge, and the lack of the methods and instruments of research, which are as yet in the future. Nevertheless, the majority of failures to find the absolute truth are due to our own personal deficiencies. It is to the correction and, if possible, removal of these deficiencies that our professional training is very rightly directed.
The naturalist should be trained in observation, experimentation, and in reasoning.
Observation is our mainstay, the foundation of all our work. I believe that in many of our laboratories a student becomes well disciplined in observation, and acquires practical acquaintance with the principal sources of error in observation in his special line of work. This part of the naturalist's education is the part best done, and we must regretfully admit that his training in experimentation is almost nil, while his training in reasoning power leaves very much to be desired.
I should like to plead before you for experimentation. It is a most difficult art—far more difficult than that of observation, because the possibilities of error are far greater. The observer inquires "What?" the experimenter "Why?" The experimenter endeavors to determine an effect and a cause. He seeks, if you will allow me the expression, to find two "whats" and their mutual relation. Every science begins with observation, and, when it is advanced enough, takes to experiments. Natural history is still in the descriptive stage. The statement is almost strictly true of meteorology and zoölogy, nearly so of geology, least so of botany. I attribute so great value to experimental work that I regard botany as being at the present time the most valuable of the natural-history sciences from an educational point of view. As regards the zoölogists—with whom I must be counted—we are most of us either systematists or morphologists. Such experimental physiological work as has been done stands not to the credit of zoologists, but almost entirely to that of medical men. In the slow advance of experimental morphology, through the labors of Driesch, Hertwig, Morgan, Roux, Whitman, Wilson, and others, we have the initiation of a most significant and beneficent reform. In all natural-history departments the great work of the future will, I believe, be done by experimenters.
For this reason it is to be desired earnestly that all young naturalists should be disciplined in making experiments. When that is done we shall hear less phylogenetic speculation and more
