ted when the plumule had begun to push through the sheath gave 5 smutty plants; finally, 150 infected when the plumule had pushed through the sheath about 1 cm., remained entirely free from smut. Microscopic examinations made a few days after the conidia were sprayed on the seedlings showed that germtube penetrations were very common in that experiment which yielded over 70 per cent of smutty plants, infrequent in those which yielded only a small per cent of smutty plants, and altogether absent in the plants which remained entirely free from smut. As in oats, the smut was confined exclusively to the panicle, and the bulk of the infections took place during the earliest stage of germination, the tissues of the growing seedling very soon becoming immune.

The results with maize were very surprising since they developed three wholly unexpected facts, viz.: (1) The germtubes are capable of penetrating any young rapidly growing part of the plant; (2) The growth of the fungous hypha which has gained entrance into the plant is narrowly localized, the sporebeds developing in situ; (3) There is no period of rest, the smut beds developing immediately, i. e. within two or three weeks of the date of infection. Previous to these experiments it was supposed that corn smut entered the plant when it was a seedling and followed the same law of development as oat smut. In the first series of experiments, which proceeded upon this supposition, the smut conidia were sprayed upon 200 seedlings in the earliest stage of germination; upon 100 which were a little older; upon 100 still further advanced; and, finally, upon 100 when the plumule was pushing through the sheath. This work was done in the laboratory and after 14 days the plants were set out in the garden. Contrary to all expectation, very few penetrations could be found even by the most careful microscopic examinations, and these were confined to the root node, none being found upon the sheath,-everywhere over the surface crept the germtubes without being able to enter. These plants were under daily observation and after 10 to 14 days a few lagged behind the rest in growth, and on being pulled up smut pustules were found on the axis a little above the root node. Of the whole 500 seedlings, only a few became smutty, viz., 4 per cent in the youngest and 1 to 2 per cent in the older seedlings. In all of them the smut pustules appeared on exactly the spot where the germtubes had entered the plant and within three weeks of the date of infection. All the other plants grew to maturity and remained free from smut. Similar results were obtained from an experiment in which soaked, ungerminated kernels of corn were planted in a dunged soil which had been

abundantly infected with smut conidia. Of the 50 plants thus treated one died at the end of 4 weeks from a smut pustule on the axis, and the rest developed without any appearance of smut. Another experiment was undertaken with 150 seedlings still further advanced, the conidia being sprayed upon them, but this also gave negative results. No germtube penetrations could be found and no smut appeared upon any of the plants. These results led to a good deal of speculation and finally to the following experiments: The first of these was with plants a foot high, having a well developed cornucopia-like summit formed by the closely wrapped bases of the large outer leaves. One hundred plants were selected and into these cornucopias a nutrient solution containing smut conidia was injected. They were covered with straw matting five days to keep off rain and then freely exposed. On the tenth day, as growth continued and the infected parts were pushed up into sight, there was a changed appearance. The parts of the leaves touched by the infectious fluid were paler than the upper noninfected parts and suggested chlorosis. This appearance was visible in different degrees on all the infected plants? Already there were slight appearances of pustules and within a day or two they became very distinct, finally covering the whole infected surface with a smutty crust. Scarcely one of the male inflorescences escaped and the axis between the leaves was also smutty in so far as the infective material could reach it. Not one of the hundred plants escaped infection, the youngest suffering most. For the next experiment younger plants were selected, i. e. those about six inches inches high. In many of these the cornucopia was not well developed and allowed the infectious fluid to run out and waste and the infection miscarried. All, however, that were large enough to retain the conidia were killed outright by the development of smut pustules, the plants twisting and curving in all sorts of shapes and frequently wilting before the smut spores were mature. The third experiment was with plants 14 feet high. Here the cornucopias were wide open and took in large quantities of the infectious fluid, which penetrated deep into the heart of the plant. After three weeks the male inflorescences appeared, but in only six plants out of 50 could any symptoms of smut be found and upon these the pustules were small and scattering. On the leaves there were wrinkled, white spots which, however, did not develop into smut pustules but subsequently became green and nearly normal in appearance. Scattered smut pustules were found on the axis at the base of the internodes in 7 cases, and the effect of the fungus was also visible on some of the upper blossoms which remained white and dried up without developing. Aside from these scattering

symptoms all of the plants remained sound, ripening normal ears. The fourth experiment, with still larger plants, gave wholly negative results. The heart of the plant proved immune, and normal ears developed. In another experiment female inflorescences were infected as soon as there was any indication of a forming ear, the Nahrlösung containing the conidia being injected into the narrow opening between the ligule and the axis. Smut pustules appeared in great numbers within 18 days but only on the parts which were actually reached by the injected fluid. Another experiment was made when the ears were in blossom. All the kernels became smutty and single ears reached the size of a child's head. In another experiment varying amounts of the lower part of the ear were protected from the fungous spray by wrapping them in blotting paper. In this case only the exposed kernels became smutty, showing again conclusively that the infection is purely local. The silk though much exposed to the conidial spray showed not the least trace of injury, having passed out of the meristematic stage. In still another experiment the kernels of the ear were sprayed with the smut conidia when they were more than grown. The result was wholly negative; no smut appeared. Another experiment showed that the adventive aerial roots can also be infected if sprayed in an early stage of their growth. In short, any meristematic part of the maize plant is liable to direct infection and this is made easy by the fact, which is also Dr. Brefeld's discovery, that the corn smut fungus, unlike that of oats and sorghum, is richly provided with aerial conidia, which are easily carried or blown from the soil to any part of the plant. The consequent desirability of keeping the soil of corn fields free froin smut spores, by removing and burning all smut pustules before they have ripened and shed, must be apparent to all. The corn smut spores seldom germinate in water, as is well known, and infection of the plant probably takes place only when the latter have an opportunity to germinate in the soil and produce the aerial conidia, this germination in the soil being greatly favored by the presence of dung. The volume contains VI, 98 pages of text and 5 lithographic plates, mostly colored.-ERWIN F. SMITH.

ZOOLOGY.

The Paroccipital of the Squamata and the Affinities of the Mosasauridae once more. A rejoinder to Professor E. D. Cope.--I. The paroccipital.-In 1870, Cope' designated the occipital externe, Cuvier, paroccipital, Owen with Huxley's name opisthotic, and homologized it with the squamosal of the Lacertilia and Ophidia. This opinion is held up in 1894 and in September, 1895,2 but for the name opisthotic the name paroccipital is then used. On the other side, it is admitted by everybody else that the paroccipital, Owen (opisthotic, Huxley), which is free in the Testudines, is united with the exoccipital in the Lacertilia; the posterior portion of this bone, which is visible from behind, has been called the paroccipital process; in its anterior portion where it reaches the basioccipital it contains the posterior semicircular canals. I have stated in my last note (AM. NAT., Nov., 1895) that in young Sphenodons the paroccipital is free from the exoccipital exactly as in the Testudines and that Siebenrock has proved without question that the outer portion of the exoccipital of the Lacertilia, which lodges anteriorly the posterior semicircular canals, represents the same element. The paroccipital process of the exoccipital in Sphenodon is, of course, identical with the paroccipital process in the Lacertilia.

To this, Prof. Cope replies: "Baur asserts that the socalled parotic process [I said paroccipital process] of the exoccipital which supports the quadrate in the Squamata is the same element as that termed opisthotic by Huxley. This I deny, and believe that in this it is Baur and not myself who has fallen into error. Siebenrock, instead of asserting this to be the case, denies it in the following language:†It is not the processus paroticus of the pleuroccipital (exoccipital) which is homologous with the (paroccipital, Owen), opisthothic Huxley, but the portion anterior to the foramen nervi hypoglossi superius which protects the organ of hearing.' Siebenrock here uses the names of Owen and Huxley as refering to the same element, but he makes the clear 1Cope, E. D. On the Homologies of the Opisthotic Bone, Amer. Asso. Adv. Sc., XIX.

Cope, E. D. On the Homologies of the posterior cranial arches in the Reptilia, Trans. Am. Philos. Soc., Vol. XVII, Apr. 27, 1892; also Am. Nat., May, 1892. The Osteology of the Lacertilia, Proc. Am. Philos. Soc., Vol. XXX, May 10, 1892, pp. 185-211. Amer. Nat., Sept. 1895, p. 855-856.

† Italics are mine.

distinction which is the important point, between the parotic process of the exoccipital and the element which contains the posterior semicircular canal. What then is the element which articulates with the quadrate in the different orders of the Reptilia?"

The sentence quoted from Siebenrock is misleading. Siebenrock does not distinguish between the parotic process of the exoccipital and the element which contains the posterior semicircular canal. He says: not only the parotic process but the whole portion anterior to the foramen is homologous to the paroccipital. This whole portion, of course, contains also the parotic process. The sentence of Siebenrock translated by Cope is printed at the end of the paper in a résumé. A full account of the conditions is given on p. 209. "Die bisherige Anschauung, dass am Processus paroticus des Pleuroccipitale (exoccipitale) das Opisthoticum zu finden sei, ist daher absolut unrichtig, sondern der ganze vordere Theil des Pleuroccipitale, welche die hintere Partie des Gehæres enthält, sammt dem Processus paroticus ist als das eigentliche Paroccipitale aufzufassen. Vergleicht man dasselbe mit dem bei den Schildkröten zeitlebens separirten Paroccipitale, so ergiebt sich schon aus der Lage und Function die Homologie der beiden Knochen." And later: "Die gleichen Verhältnisse bestehen bei Hatteria, nur bleibt hei dersellben das Paroccipitale viel laenger vom Pleuroccipitale (exoccipitale) getrennt, als bei Lacerta."

That Prof.Cope has not studied Siebenrock's paper is also evident from the following sentence: "In the Testudinata, and according to Baur, in Sphenodon, the element which extends externally from the exoccipital to the quadrate is continuous with the opisthotic, but the semicircular canal is included in its proximal part only. Here the structure is entirely different from that which characterizes the Squamata, where the opisthotic does not extend distal of the canal and fuses early with the exoccipital." It is still more evident from the following words: "In the Squamata, where the opisthotic is restricted to the region of the canal and does not reach the quadrate, this socalled paroccipital is distinct." Cope thinks the paroccipitalotic portion of the paroccipital or opisthotic in the Testudines is not homologue to the paroccipital +otic portion of the paroccipital or opisthotic of the Squamata, and has the idea that this bone, paroccipital, Owen, opisthotic, Huxley, occipitale externe, Cuvier, consists of two elements, the outer one-the paroccipital-and the auditory portion, the opisthotic. He admits that "the direct evidence for such a primitive division of this element (occipital externe, Cuvier; paroccipital, Owen; opisthotic, Huxley) † Italics are mine.