Summary This summary of the physical and chemical properties of wood shows that a large amount of experimental work must be done before wood can be completely characterized, and the interrelationship of all of its properties understood. The situation looks hopeful, however, when it is observed that practically all of the physical information which is available, and which has already helped greatly in the solution of the interrelated problems has been collected very recently (see dates on references given). It seems reasonable to suppose that further experimentation along physical lines will be one of the chief factors if not the chief factor in solving the problems of the characterization of wood. Here as in all other fields of scientific endeavor, however, care must be taken not to lay aside any of the old tools or methods of procedure merely because others for the time being may be more productive of new scientific information.
Gem-Killing Powers of Smoke Studied. Determination of the efficacy of smoke as a germ killer is the object of a series ol experiments recently reported by Dr. Ernest Hess of the Biological Board of Canada t o the Society of American Bacteriologists a t Rochester, N. Y. Increases in the amount of air admitted per hour to the smoke producer caused uniformly a greater reduction of bacteria as well as "thicker" smoke. Smoke from different types of wood showed practically no difference in bactericidal power after six hours' adion upon the test organisms, Dr. Hess stated. The tests were undertaken t o get some sort of check on the value of smoke-curing as a means of preserving fish. Dr. David I,. Belding of Boston University reported the isolation of the causative organism of a blood infection of fishes that has caused serious losses among the fish of New England and Middle Atlantic States. The infection seems peculiar t o eoldhlwded animals. While no satisfactory preventive treatment has been discowed, some epidemics have been held under control in hatcheries by keeping the water below 55 degrees Fahrenheit.-Science Servirc Tests Fail to Predict Students' Grades. College students must continue t o submit to the painful process of taking examinations in order that their teachers may determine their achievement. Experiments t o predict what they would do in psycho lo^ courses by psychological tests given in advance have not met with success. Before the section of psychology of the American Association for the Advancement of Science. Dr. Joseph Peterson, of the George Peabody College for Teachers, told of his researches. The experiments were made with two hundred sophomore students, with either one, two, or three different tests, but practically no correlation was found between the success in the tests and the students' final grades. Professor Peterson thinks that the diverse interests and social activities of the students account for this failure of the tests in prediction. I n other words, he says, "other things enouah . than intelligence (dven . . intelligence for one to become a sophomore college student) have important rbles in determining success in psychology classes as measured by the examinations. This is roba ably true also for any other college subject."-Science Service ~