The Germ Pore. The relation of the type of stamen and potency of pollen to the germ pore should not be overlooked. The relation between the reflexed stamen and impotent pollen has already been emphasized. It is sufficient now to merely mention the fact that in the impotent pollen, such as that formed by Brighton and Barry, or in the pistillate flower of the wild Vitis vulpina, no germ pore is formed. Hence, we have the relation-smaller cells on the outer surface of the filament, reflexed stamen, impotent pollen, no germ pore. This relation, of course, is found typically in those forms which are completely self-sterile.

Relation of Sterility to Other Factors. Earlier in the paper the importance of the relation of sterility to hybridity, flower type and dioeciousness, was emphasized. The relation between impotent pollen and flower type has already been pointed out, but these two factors have not heretofore been emphasized in connection with the absence of the germ pore in the impotent pollen. The relation of the reflex type of stamen to dioeciousness in the grape is well known. It now remains to be seen whether there is a relation between sterility in the grape and hybridity. A number of hybrids between species are self-sterile, but in the case of the grape, this condition does not necessarily seem to be important, since we have both sterile and fertile hybrids between species. It may also be mentioned that we have varieties with either upright or reflexed stamens, as the case may be, which are hybrids between species.

Cases are recorded where sterility is thought to result from the bringing together in the zygote of an unequal number of chromosomes from the parents. This does not seem to be the case of Brighton or Barry, which are two Labrusca Vinifera hybrids. The reduced number of chromosomes in Brighton, Barry, Concord, V. vulpina and V. vinifera is twenty. Brighton is a cross between Concord and Diana Hamburg. Concord, the pollen parent, is generally considered to be a variety of V. labrusca, and Diana Hamburg, a cross between V. labrusca and V. vinifera. It will be seen, then, that since Brighton has an equal number of chromosomes from both parents, an unequal number in this case at least does not seem to be a factor in sterility.

Discussion. Sterility resulting from difference in essential or gans should be carefully distinguished from that due to unfavorable weather conditions which occur at flowering time during many seasons. The latter influence comes only occasionally while the former is present irrespective of seasons and should be taken into consideration at the time vineyards are set.

We have seen that sterility is associated with both hybridity and dioeciousness. The native species of our grapes are, generally speaking, dioecious, and fertile pollen borne by the pistillate flower with reflexed stamen, is rare. Hence, this condition suggests a deep-seated relation between sterility and functional decliny. Sterility cannot necessarily be attributed to hybridity since we have both fertile and sterile hybrids. Hence, taking into consideration this fact, it seems that the contributory causes of sterility are deepseated influences operating to produce declinism or dioeciousness.

Considering only the wild types of Vitis, which are essentially dioecious, we have seen (Dorsey, '12) that intermediate forms do occur between the pistillate flower with reflexed stamen and the staminate flower with upright stamens. This condition might be interpreted in accord with Booth's contention ('02) that "the grape is in a state of evolution from an older hermaphrodite form to forms that are essentially staminate and pistillate." The cytological evidences presented in this paper, however, is not to be interpreted to affect the question of the evolutionary development of Vitis.

Conclusion. When the. question of sterility in the grape was first given careful attention by practical men, the question was continually recurring as to whether it was of such a nature that it could be overecome by cultural conditions. The deep-seated nature of sterility and its relation with dioeciousness would seem to indicate that a negative answer must ben given to this question. Beach, S. A. Self-Fertility of the Grape. N. Y. State Expt. Sta. Bul. 157:397-441. 1898.

Fertilizing Self-Sterile Grapes. N. Y. State Expt. Sta. Bul. 169:331-371. 1899.

Investigations Concerning the Self-Fertility of the Grape. 19001902. N. Y. State Expt. Sta. Bul. 223:269-290. 1902.

Booth, N. O. (a) A Study of Grape Pollen. N. Y. State Bul. 224:291-302. Pls. I-VI. Figs. 1-24. 1912.

Dorsey, M. J. Variation in the Floral Structure of Vitis. Bull. Torr. Bot. Club. 39:37-52. Figs. 1-24. 1912.

Munson, T. V. Investigation and Improvement of American Grapes. Texas Expt. Sta. Bul. 56:217-285. Pls. 1-19. 1899. Foundations of American Grape Culture. Dennison, Texas. 252 pp. Pls. 89. Fig. 13. 1909.

Reimer, F. C. and Detjen, L. R. Self-Sterility of the Scuppernong and Other Muscadine Grapes. N. C. Expt. Sta. Bul. 209:5-23. Fig. 13.

1910.

THE EFFECT OF POLLEN OF WAGENER AND MCINTOSH ON THE
SIZE OF AND NUMBER OF SEEDS IN WEALTHY APPLES.
By J. W. CROW, Ontario Agricultural College, Guelph, Canada.

The following table sets forth results secured this year (1913) in making certain crosses with the Wealthy apple. The writer presents the facts of the case for the purpose of recording the same, and to the end that accumulation of evidence may shed light on some of the more intricate phases of pollen influence. The results herein presented are in harmony with a result secured some years ago in making certain experiments into the amount of pollination required in apples. The quantity of pollen was regulated by removing one or more of the five stigmas which normally occur. It was found at that time that the largest number of good seeds per fruit were obtained where only three stigmas were pollenized, or, in other words, where only a three-fifths quantity of pollen was used.

The best setting of fruit and the largest number of mature apples, however, were found to result from complete pollination of all five parts of the pistil. The inference is, therefore, that number of seeds is not always a reliable index of the amount of pollination which has taken place. It appears, too, that pollen in excess of the amount required to accomplish maximum fertilization of seeds, fails to increase proportionately the number of good seeds per fruit, but does increase the number of fruits which set and which mature. In the present instance, the McIntosh pollen is less effective in producing good seeds, but has decidedly more effect than the Wagener pollen in increasing the size and number of apples.

These tables show: (1) that in this case, McIntosh pollen is more effective in producing a set of fruit on Wealthy than is Wagener pollen. (2) The advantage gained by McIntosh pollen is approximately maintained throughout the season, and the resulting fruit is more abundant in proportion to the number of blossoms pollenized. (3) The average size of the fruits is larger under the McIntosh pollen. It should be stated here that all the fruits in question were produced on the same tree and that there was no advantage to the fruit of either lot in the matter of location or distribution. The color of the apples produced by McIntosh pollen was noticeably darker and richer, but there is unfortunately no method of measuring the amount of difference. (4) The average number

of good seeds per fruit is greater under the Wagener pollen, in spite of the fact that the average number of seeds which had undergone at least partial development is approximately the same in both cases. There is, thus, in this case, an inverse ratio between the number of good seeds and both the number and size of the apples of the two lots. It is evident that the McIntosh pollen has had some effect on fruit setting and the size of fruit entirely separate and distinct from its effect in bringing about actual fertilization. RASPBERRY BREEDING.*

BY RICHARD WELLINGTON,

Experiment Station, University Farm, St. Paul, Minn.

The breeding of raspberries is not a new subject,, as is shown. by the multiplicity of varieties, nor has it received the attention that it deserves by the modern plant breeders. Why has this important economic plant been neglected? It contains many diverse forms, propagates and crosses readily. Perhaps we have too many annuals that need investigating and naturally an investigator selects the subject which will give him the quickest results, especially in this period when new genetical knowledge is held at a premium. If, however, one commences breeding with the Rubus neglectus, and its parents, Rubus strigosus and Rubus occidentalis, this objection is partially alleviated, since segregation of characters takes place imemdiately in the first generation of the selfed hybrid. This method was pursued at the New York Agricultural Experiment Station and consequently it is possible at this early date to give results in work started in 1910 and 1911. The data at hand are too meagre to draw any marked conclusions; and yet its presentation is deemed excusable, if it will either contribute to the present knowledge of raspberry breeding or act as a stimulent to further experimental work.

In crossing varieties, emaculation of the mother is requisite, as well as the proper knowledge of the time of performing the same. Paul J. Anderson, formerly of Cornell University, found in his microscopic examination of unopened red raspberry blossoms that, not only the pollen was on the stigmas, but in many cases the pollen tubes had commenced to penetrate the pistils. This observation caused the writer to make this past season a careful study of the proper time for emasculating, for, if fertilization takes place in the bud stage, then the thousands of crosses at the New York Agricultural Experiment Station have no scientific value. Instead of depending upon microscopic examination, buds of different stages of development were selected and emasculated. The results of this work are presented in the following table:

*The material used in this experiment was grown at the New York Agricultruar Experiment Station, Geneva, New York. The writer gives thanks to Dr. U. P. Hedrick for his suggestions and the use of the data in this article, to Dr. M. J. Dorsey, W. H. Alderman and G. II. Howe for their assistance in crossing, and to O. M Taylor et al. for their care in the propagation of the plants.

From the study of Table I, one can safely conclude, at least for the varieties tested; first, there is no danger of self-fertilization taking place if the buds are emasculated about one day before the time the calyx commences to split; and second, the buds examined by Anderson must have been either of another variety or in the advanced stages. It should be remembered, further, that Anderson's work was in regard to the effect of cement dust on the setting of fruits and, therefore, the exact time for emasculating did not come under his consideration. Split buds, however, should not cause an error in breeding work, for a careful investigator would readily note that they not only possess protruding pistils and usually one or more bursted anthers, but are frequently visited by insects. As a universal rule, it is safe that, if the anthers in a flower have not burst and the pistils have not been exposed, crossing or selfing is an impossibility.

In 1910 many blossoms of the Columbian, a purple raspberry, were self-fertilized, the seed was stratified out-of-doors soon after harvesting and was sown early the following spring (1911) in flats kept in the greenhouse. As soon as the young plants obtained sufficient size, they were pricked out in other flats, hardened off in due time and finally transplanted early in the summer-113 in toto-between the vines of a young vineyard. The plants flourished at first but when the first winter overtook them several fell by the wayside, a few were killed to the ground and later recovered, a few were partially winter killed, and many passed unharmed. Several plants bore a partial crop in 1912, but since many plants had no fruit, the final records could not be completed until this year. The main characters observed and recorded were the color of causes and fruit, amount of spininess and glaucousness, hardiness, date of leaf bud breaking and fruit ripening, and freak blossoms and fruits.