The experiments of Diirken (1923) on the effect of coloured lights on the larvae and pupae of the cabbage butterfly, which seem to him to prove the inheritance of an acquired character (colour changes), while furnishing quantitative data are capable of a different interpretation, namely, the selection of those indi viduals for parents that had already an inherited difference of response to light. A careful series of experiments by Sumner on the effect of temperature in the length of foot and tail in mice and on the possible transmission of the effect are probably significant, but the results are not claimed by him as demonstrating the in heritance by means of a changed somatic character. H. Przibram, experimenting in carefully controlled conditions, found that the temperature at which rats are kept has an effect upon their tail length. He further found that when parents were kept at extreme temperatures, this did have an effect upon the tail length of the young, even when these were reared at normal temperature. However, when the temperature of the parents' environment was very extreme, the tail-length of the young was altered in the opposite direction from what would be expected on Lamarckian principles; and secondly, the effect was not transmitted to later generations. Now the experiments, while opening up interesting lines of research, are wholly contrary to Lamarckian views. The recent announcement of Pavlov, that a conditioned reflex, estab lished in mice, is inherited and shows marked advances in each generation, has been withdrawn by Pavlov as due to an error. The results as announced were, moreover, in direct contradiction to somewhat similar experiments by Bagg, McDowell, and Vicari.
There is another class of experiments carried out in recent years that has also opened up a wide field for discussion, namely, the direct and indirect effect of the external environment on the germ-cells themselves. There are two distinct possibilities involved in the situation. First, it has been claimed that the same kind of effect produced in the body-cells by an external agent may simul taneously affect the germ-cells (egg or sperm) in such a way that they produce a new individual showing the same trait (parallel in duction). Second, that the external agent may produce a change only in the germ-cells. which change appears in the individuals of the next generation developing from such germ-cells (direct induc tion). The same effects may then appear in later generations, because the affected germ-material itself is handed on.
The first supposition, however plausible it may sound, may rest on a false analogy. The response of the body cells to an external agent is presumably a reaction depending on their spe cific differentiation. This view is supported by the absence of a similar response in other kinds of organs or tissues. What grounds, then, are there for supposing that the undifferentiated germ-cells should be affected specifically but not show their response until the definite organ in question develops? There can be given no rational explanation for such an expectation, which obviously rests on an assumption at variance with what is known to physiology.
There is no positive evidence to appeal to in support of this sup position, although there are a few cases which may at first sight seem to come under this heading. For example, Guyer injected crushed lenses from the eye of a rabbit into fowls with the inten tion of producing in the blood of the fowl an antibody. Later some of the blood of the fowl was injected into a pregnant rab bit. In rare instances one or more of the young rabbits that were born showed abnormalities of some kind in their eyes. From a few of those affected individuals offspring were produced which sometimes had abnormal eyes, but it was not claimed that the same specific abnormalities always appeared in parent and off spring. Several explanations may be offered. It is known that occasionally rabbits are born with defective eyes, but the absence of abnormal eyes in the control, unoperated relatives of those used, was supposed to meet this argument. Until the frequency of the occurrence is known the reply is not entirely adequate. Never theless, the facts as reported appear to make it improbable that the connection was accidental. J. S. Huxley and A. M. Carr Saunders have repeated Guyer's experiments and failed to produce inherited defects in the eyes of the offspring, and other failures in mice and other species have been reported by other workers. Guyer has also later reported that when the lenses of rabbits are destroyed by a needle, a few of the offspring may have eye defects. This result would seem to mean that a parent with injured eyes may produce something (antibody?) that affects the eyes of the developing young in utero, but whether the germ-cells are af fected has not yet been stated, much less proven. A more surpris ing instance of the apparent inheritance of an acquired character is reported by Griffith. Adult rats were rotated in cages for several months. They became disequilibrated, showing a tendency, when removed, to circle in the direction in which they had been ro tated. Later, pus was discharged from the ears and they did not live long. Some of the offspring (6o out of 500 born from parents that had been previously rotated) also showed a tendency to circle and a preponderating number in the same direction as their parents. Detlefsen has reported that rats with diseased or defec tive ears occur in a small percentage of cases. They show the type of behaviour characteristic of Griffith's rats. Whether a diseased condition of the parent ears could lead to infection of the ears of the young rats has not been tested, and even if it were, the specific inheritance (duration of rotation) described by Griffith remains to be explained.