C. picta when blindfolded usually rushed off a surface at any height without the least hesitation. There is no evidence, from my experiments, that the tactual and muscular impressions received when the legs are stretched over the edge have any inhibitory influence on the movement. From this it is clear that the hesitation of this species observed at heights of 180 cm. is due to visual impressions, not to the unusual organic impressions received. This species at first tries to remove the covering from the eyes by rubbing the fore legs over the head, but failing it soon becomes accustomed to the blindfolded condition. N. guttata is much disturbed by the obstruction of its vision, and for long periods persistently tries to remove the cap. Most individuals after a time move about freely, but whenever they reach the edge of the board they turn back. Evidently the tactual and muscular impressions inhibit the tendency to move forward. Whereas in case of C. picta, we see the blindfolded animal risking falls which it would not have risked in its normal condition, in N. guttata we see exactly the reverse, for as a rule the animal when blindfolded does not leave the board. T. carolina does not struggle so persistently to remove the covering as do the other species, but it is inactive when blindfolded. It behaves in general much as it does when placed at a height of 180 cm. above the floor. This indicates that it depends upon vision for guidance in its movements to such an extent that it is not likely to move about much unless it can see clearly. Visual impressions are of prime importance in the space perception of tortoises, and tactual, muscular and organic data occupy a position of secondary importance. Yet there are many reasons for believing that we often underestimate the value in the reactions of simple organisms of that complex mass of sense impressions which we are not as yet able to refer to specific organs. JAMES ('90, II, p. 150) has called attention to a fact that is significant in this connection; "Rightness and leftness," "upness and downness," he says, "are again pure sensations differing specifically from each other, and generically from everything else." We are inclined to lose sight of the organic impressions, and to refer reactions to data received through the so-called special senses. Many experiments have already been made which show that the direction of turning, apart from vision, is extremely important in the motor habits of tortoises. and frogs. I gratefully acknowledge my indebtedness to Mr. SAMUEL HENSHAW for suggesting to me the desirability of a comparative study of the space reactions of tortoises; to Mr. THOMAS BARBOUR for valuable assistance in many ways, and for the opportunity of observing the behavior of several foreign species; to Mr. Wм. T. HORNADAY, director of the New York Zoological Park, and to Mr. R. L. DITMARS, Curator of Reptiles, for the privilege of conducting experiments in the Park, and for many courtesies, and to Mr. C. W. HAHN for the use of tortoises which were in his possession. James, Wm. '90. REFERENCES. Principles of Psychology. New York. Mills, Wesley. '98. The Nature and Development of Animal Intelligence. London, 307 pp. Small, W. S. '99. Notes on the Psychic Development of the Young White Rat. Amer. Jour. Psychology, Vol. 11, pp. 80-100. Thorndike, E. L. '99. The Instinctive Reactions of Young Chicks. Psychological Review, Vol. 6, pp. 282-291. Watson, John B. '03. Animal Education. Chicago, University of Chicago Press. 106 pp. A NOTE ON THE SIGNIFICANCE OF THE FORM AND CONTENTS OF THE NUCLEUS IN THE By SHINKISHI HATAI. (From the Neurological Laboratory of The University of Chicago.) With Plates III and IV. In the course of an investigation on the growth changes in the nerve cells of the white rat, the writer noticed that the nucleus very often had a peculiar shape; the shape being similar to an amoeba exhibiting pseudopodia-like processes along one side. Careful observation revealed the fact that this peculiar form of the nucleus occurs normally during the period of the early growth and in attempting to explain it we meet several interesting questions of histological as well as physiological importance. The results here reported form a part of a series of observations on the growth changes in the developing nerve cells. For this investigation, the intrauterine embryos of cat, pig and white rat were used. The present description and drawings are, however, based entirely on preparations from white rat. Unless otherwise mentioned, the description also applies to the cat and pig. The tissue was preserved in a normal salt solution saturated with HgCl; in CARNOY'S SOlution and in GRAF'S chrom-oxalic mixture. For the microchemical test for P. and Fe., however, the tissue was fixed advantageously with 95% alcohol. The paraffine sections were made 3 to 6 micra in thickness, and were stained with HEIDENHAIN'S iron-haematoxylin alone or sometimes followed by 1% aqueous solution of eosin and BIONDI-EHRLICH'S tri-color stain. Other sections were stained with toluidin blue and eosin. General characters of the spinal ganglion cells.—The spinal ganglion cells in embryos of the white rat, 10 to 13 mm. long, present a bipolar shape in most cases; one of the processes contains a large amount of the cytoplasm and is recognized easily because it stains more deeply than the other. The second process arises from the opposite side of the cell-body and stains faintly. It contains a very small amount of cytoplasm and is hard to distinguish from the surrounding structures. Only the former process is shown in the figures. The relation of these processes to the spinal cord will be described in a future paper. Hereafter, in this paper, the term process means the former branch, rich in cytoplasm. The nucleus is very large compared with the cell-body and presents a more or less oval shape. It contains a large number of the minute granules among which two different forms may be distinguished, not only by their size but also by their staining reactions with iron-haematoxylin; one form of the granules stains deep black while the other presents a grey tint. The granules which stain an intense black are very much larger in size and occur most abundantly along the nuclear wall and its vicinity; the faintly staining granules, on the other hand, are very small in size and appear to form a fine network in the nucleus. This network is most condensed around the larger granules of the former group. The large granules are identified with the basophile substance and the small granules with linin or oxyphile substance. This grouping is by no means satisfactory, for by using the BIONDI-EHRLICH stain, as well as by applying the microchemical tests, it has been noticed that among basophile granules (larger form) there are several different kinds which stain with different intensities and similarly there are several different kinds of the oxyphile granules. So far, therefore, as color reactions are concerned, the oxyphile and basophile granules grade into one another and no sharp distinction can be drawn between the two. This fact is extremely important in connection with the present investigation and will be discussed more fully later on. Among the large granules, sometimes one and in some cases more than one, can be distinguished as exclusively composed of the basophile substance, but in many cases, the large granules contain both basophile and oxyphile substances. When this occurs and the basophile surrounds the oxyphile substance then such a granule may be regarded as a nucleolus of the adult nerve cells. Enlarged granules of this nucleolar type are shown in Figs. 1 and 2. The position of these granules is not constant but they lie in some cases along the nuclear membrane and in others they occupy the center of the nucleus (Fig. 1). Shape of nucleus.—Changes in the shape of the nucleus and the alterations of its position in the cell have been noted by a number of observers. The phenomena have been observed especially under experimental and pathological conditions. In the normal condition, however, they have been reported by only a few investigators. Several investigators observed a pocket formation along the nuclear surface in the spinal ganglion cells of fish. These invaginations are repeated several times in one nucleus, some of them being deeper than the others, and thus the nucleus presents pseudopodia-like proSuch an appearance is rather common in the nuclei of the nerve cells in spinal ganglia and ventral horn of lower vertebrates (HOLMGREN), but on the other hand, it is rarely visible in the cells of the higher mammalia, and when it occurs it is not so conspicuous as in the lower forms. The present writer had the opportunity to examine a large number of the preparations of the nerve cells of the white rat at different ages, but failed to find the pseudopodia-like processes of the nuclei in animals of one day or older. As a rule, after the age of one day the shape of the nuclei is constantly ovoid or spherical and does not show pseudopodia-like processes. cesses. By examining the nuclei of spinal ganglion cells of embryos (10 to 13 mm.), the following appearances have been observed: As is shown in the figures, the nucleus of the embryonic spinal ganglion cells lies, as a rule, at one side of the cell body; that is it lies eccentrically. Such an eccentric location of the |