lieving that "absorptive glands" exist in the leaves of Drosera rotundifolia and Pinguicula vulgaris, and that bodies more or less analogous to these glands are found in all the other plants which have to the present time been regarded as carnivorous. Thus, they exist in Nepenthes and Dionaea, and the "quadrifids" of Utricularia strongly resemble them. The interesting fact is also stated by Mr Bennet that, with the exception of Callitriche verna, these absorptive glands do not occur in the leaves of any other than insectivorous plants. Though further researches on this subject are desirable, yet the above observations by so able and competent a botanist strongly confirm the views just stated, and point to the provision of special organs by which the process of absorption of animal food is accomplished.

We now proceed to consider the third great question -viz., that of assimilation. It is the part of the process to which probably most objections have been taken, even Professor Morren refusing to commit himself at once to it. I mentioned in my previous paper a fact communicated to me by Mr Lindsay, a very careful observer, that Dionaeas covered with glass jars did not thrive so well as those which were fully exposed. This seemed to me to prove satisfactorily that the absence of insects was prejudicial to the healthy life of this plant. We determined, however, to put this to the test of fresh experiment. We had plants placed under large wide-mouthed jars, one of which was left open, while the other two were so arranged that water and air were freely admitted, but no insect could obtain access to the enclosed plants. (I show you the apparatus employed). For months they were left in the same greenhouse, and no difference was apparent in the general appearance and growth of the two sets of plants; and even at the end of two or three years the plants excluded from insects appeared as healthy and strong as the others. I always felt, however, that mere appearance was not a sufficient guide in a matter of this kind, and hence I never communicated the results to this Society. Besides, when I considered the wonderful structure of the Dionaa, and the admirable adaptation of all its parts for their allotted functions, I could not but regard it as highly im

TRANS. BOT. SOC. VOL. XIII.

2 B

probable that such complicated contrivances were not designed for the nourishment of the organism. Casimir de Candolle has more recently asserted that the opinion regarding digestion and absorption in the case of Dionæa can only be regarded as an hypothesis until direct experiments have proved that the substances so digested and absorbed are assimilated-i.e., are utilised for the formation. of new tissue. To test this De Candolle performed experiments similar to my own. One set of plants were fed with pieces of meat and white of egg, as well as insects, while the other two plants had only what they got from the soil. After six weeks' careful study of them he could detect no difference in their general appearance, or in the number and size of the new leaves which had been produced during the experiment. By many direct measurements he was satisfied that the different parts of the same leaf did not increase through animal food being digested. He owns that his experiments were too few, and I do not know that he has ever carried out his resolution to prosecute the subject further.

The fallacy of these experiments has, however, very recently been shown by Mr Francis Darwin, who, by most careful observations on Drosera, has proved that assimilation does undoubtedly occur, though, as its more marked indication is to be found in the organs of reproduction, it is easy to understand how other observers have been led into error. As I am not aware of any very exact experiments having been made on Dionaea in proof of assimilation, I shall refer to those where Drosera was chosen as the subject of experiment, and considering the many essential points of resemblance between that plant and Dionaea, we may feel justified in attributing like qualities to the latter.

These experiments of Mr Francis Darwin were conducted with great care, all likely sources of fallacy having been avoided, so that I need not detail the plan followed farther than to say that the soup plates in which the plants were placed were divided into two equal portions, on the one side of which were the unfed plants and on the other the fed ones. Roast meat was used as the diet of these last,

of a grain being placed on the secreting glands. The

experiments began on June 11th of last year, and ended on August 25 of same year. The first difference observed was on July 17, when the fed side, taken as a whole, was decidedly greener than the unfed side. This change was manifest in all the plates. The tentacles on the unfed side were of a redder colour than those of the fed plants, hence chlorophyll was more developed in the latter; and it was proved, by drying and weighing, that a much greater amount of cellulose was formed by the fed than by the unfed plants; because chlorophyll is associated with an increased assimilation of carbonic acid, and thus permits more cellulose to be formed. An average leaf from fed and unfed sides was examined on July 18, and a marked difference was observed both by the naked eye and by the microscope; in the former case it was evidenced by the dark purple hue of the fed leaf, and in the latter by the large and numerous chlorophyll grains crowded with starch, with which its cells were filled. He attributes the absence of appreciable increase of starch on August 16, 17, and 21, either to its having migrated to the root-stocks and flowering stems, or to its having been used up in larger quantity for the formation of the greater amount of cellulose, which we have seen was found.

To these experiments we shall immediately return; meanwhile we give his final results, which are that the carbohydrates are formed in far greater amount in the fed plants. The body of the chlorophyll grain, he adds, is protoplasmic, and hence an increase of nitrogen will favour formation of chlorophyll, and increase the starch-producing powers of the plant. Fraustadt, however, had come to an opposite conclusion regarding the starch, for he had found that it diminished in Drosera with absorption of organic matter by the leaves. Darwin's reply is simply-" This result may be perhaps attributed to over-feeding;" but this is by no means satisfactory, for the plant, which had caught an insect, on which he made his first and third experiments, was found deficient in starch. In the first instance the transverse sections through the middle of the petiole, where he found only about five or six cells containing starch, and these lying in the neighbourhood of the middle, largest vascular bundles. In the third experiment, where the

transverse section was through the leaf, there was not a trace of starch. Fraustadt's first experiment at least seems to have been made on a leaf which had not been fed artificially, but which had naturally caught an insect, small pieces of whose skeleton were all that remained, so that he could determine no more regarding it; and as his other experiments were with leaves subjected to the same condition, it is not likely that over-feeding would occur, for it has always appeared to me that this misfortune rarely if ever occurs in a state of nature. But Darwin seems to forget that the experiments were in pairs, the second referring to those leaves which had never had organic matter, in which the contents of the cells in the petiole wholly, or for the most part, gave the distinct reaction of starch. Fraustadt's fourth experiment was a transverse section of the midrib of the same plant which was used in his second one, and he found there that the cells were very abundantly filled with starch.

Darwin attempts to explain the deficiency on August 16, 17, and 21, as above stated, by two considerations; the first of these is, that the starch had migrated to the root-stock and flowering stems. Now as to the root-stock, it is a significant fact that the little ears of the base of the petiole, which embrace the rhizome, were, according to Fraustadt, to an extrordinarily abundant degree filled with starch, and this condition exists equally in cases where the leaves have absorbed animal or more general organic food, as where they have had none, and Fraustadt's conclusion is that, since where organic matter is given the starch grains disappear only in the portions above ground where the cells contain chlorophyll, the assimilation or the production of carbohydrates by chlorophyll and the absorption of organic matter exclude each other: but, he adds, that the presence of chlorophyll does not exclude the reception of organic substances.

With such diametrically opposite opinions before us, we must conclude either that Drosera and Dionaea are differently affected by organic matter absorbed by the leaves, which would be in the face of all sound analogy, or we must accept F. Darwin's other suggestion, that the starch had gone to form the increased amount of cellulose which,

as we have already seen, he found in his fed plants. We must, however, never forget that chlorophyll is not essential to the production of starch (if not also of other carbo-hydrates), as the experiments of Van Tieghem have shown.

But let us now look at some of F. Darwin's tables, and we shall be forced to the conclusion that real nutrition of the plant, and especially of some portions of it, is effected by the absorption by the leaves of organic matter. Table 3 gives the number of flower-stems on each side in the six plates on August 7, and the proportion is seen to be, for unfed to fed, as 100: 149.1.

Table 5 again exhibits the number of healthy leaves (as shown by their having secretion on their glands) on each side of three of the six plates, and here the proportion of unfed to fed is as 100:136 9. Experiment about middle of August.

From Table 6, among other things, we learn that the total weight of unfed and fed plants (without flower-stems) dried at from 80° to 90° C. was in the proportion 100: 121.5.

In Table 7, among other items, the total number of stems (including those bearing flowers as well as those bearing capsules) from all the plates gave the proportion of unfed to fed as 100: 1649; and the weight of these was as 100:231-9; the average weight per stem being as 100: 141.3.

In Table 8 we have the total number of capsules, which gives the proportion of 100 unfed to 194-4 fed; average number of capsules for each stem, 100: 1216; total weight of capsules unfed to fed, as 100: 130; total number of seeds in capsules, as 100:122-7; average number of seeds per capsule, as 100: 122-7, same as above.

Table 9 furnishes us with the average weight of each seed, which was in proportion of 100:157-36; total number of seeds from all the plants in all the plates gave the proportion of 100: 241.5; total weight of seeds from all the plants in all the plates was in proportion of 100: 379.7.

It is very striking that the greatest disparity in the proportions is in the instance last cited; on this Mr F. Darwin remarks, that "the special advantage which nitrogenous food, given to the leaves, yields is the power of producing