(Colletotrichum gloeosporioides Penzig), a fungus which frequently attacks the young flower spikes and destroys them. Its activities are largely dependent upon weather conditions, being greatest when the air is warm and moist, with frequent showers. This question was studied by McMurran. Sometimes the flowers appear in dry weather and escape unharmed, but a few days of showery weather later on may provide favorable conditions for the development of the fungus, and the young fruits may be destroyed or badly disfigured and deformed. In some seasons trouble from this source is relatively slight, in others serious.

Allowing for injury from this cause, which is an important factor in the production of mangos, it was believed that there must be other factors responsible for many of the crop failures of the Mulgoba and other varieties. Several Indian and West-Indian writers have suggested that defective pollination might be one of the most important factors, and it was for the purpose of determining its precise importance that the investigations herein described were undertaken at Miami, Fla., in 1915 and were continued in 1916. In connection with any studies such as this on the pollination of fruitbearing trees, two papers by M. B. Waite should be examined.

THE MANGO FLOWER AND ITS POLLINATION.

It has been affirmed by Hartless that the mango is largely, if not solely, wind pollinated. Other writers have also advanced the opinion that the wind is an important agent in effecting pollination.

It seems evident, however, that the mango has none of the characteristics of an anemophilous plant, but, on the other hand, presents well-developed adaptations to insect pollination, so that it may be considered truly entomophilous. The amount of pollen which reaches the stigmas was found to be slightly greater in one instance on a very windy day; but this undoubtedly was due to the brushing of one flower against another or against the surrounding foliage, the panicles being thrashed violently about.

Some of the principal characteristics of an anemophilous plant are that (1) the pollen is abundant, compensating for the enormous waste in transport; (2) the pollen grains are dry and incoherent, so that they are easily carried by the wind; and (3) the stigmas are commonly large and bushy and freely exposed, so as to have every chance of catching the floating pollen grains. The mango has none

1 McMurran, S. M. The anthracnose of the mango in Florida. U. S. Dept. Agr. Bul. 52, 15 p., 4 fig., 4 pl. 1914.

2 Waite, M. B. The pollination of pear flowers. U. S. Dept. Agr., Div. Veg. Path. Bul. 5, 110 p., 5 fig., 12 pl. 1894.

1899.

Pollination of pomaceous fruits, In U. S. Dept. Agr. Yearbook, 1898, p. 167-180, fig. 32–44.

* Hartless, A. C. Mango crops and some factors influencing them. In Agr. Jour. India, v. 9, pt. 2, p. 141-159.

1914.

1

of these characteristics; it produces comparatively few pollen grains, often not more than 200 in an anther, and never more than 1,200 in any that were examined; and it must be remembered that this represents the total number produced by one flower, since there is commonly but one fertile stamen. The pollen grains show a decided tendency to cling together, especially in damp weather; and even on bright, sunny days it was found difficult to dislodge them from the anther by subjecting them to the full draft of an electric fan for 30 minutes, most of the grains still clinging to the anther at the end of that time. The stigma is exceedingly small and not provided with projections of any sort to assist in catching pollen.

On the other hand, the production of honey for the attraction of insects shows a distinct adaptation to insect pollination. The structure of the flower is such as to entitle it, apparently, to be placed in Müller's biological class "A", or "flowers with freely exposed honey." Flowers of this class are visited by insects of several orders, from the short-tongued Coleoptera to the longtongued Lepidoptera, and members of both these orders, as well as of Diptera and Hymenoptera, have been observed on mango flowers, as will be described later on.

STRUCTURE OF THE FLOWER.

The mango is polygamous and produces its flowers on terminal panicles varying in length from a few inches up to 2 feet, each panicle carrying from 200 or 300 up to more than 4,000 flowers, of which only 2 or 3 per cent are perfect in some instances; in others as many as 60 or 75 per cent. (Pl. I.) The character of the panicle and the number of flowers produced upon it vary with different varieties, as also the length of time they remain in bloom. Some varieties remain in flower but 10 days, others for nearly 2 months, and on one panicle of the Sandersha 4,200 flowers were counted which opened. at the rate of 20 to 240 a day, extending over a period of 40 days.

The individual flower 2 is subsessile, 6 to 8 millimeters in diameter when the corolla is outspread, the calyx composed of five ovatelanceolate, finely pubescent, concave sepals and the corolla of five elliptic lanceolate to obovate-lanceolate petals, 3 to 4 millimeters long, whitish, with three or four fleshy orange ridges toward the base, and inserted at the base of a fleshy, almost hemispherical disk, obscurely 5-lobed and usually about 2 millimeters in diameter. In the perfect flower the disk is surmounted by a globose-oblique

p. 485. Leipzig, 1881.

1 Müller, Hermann. Alpenblumen
Knuth, P. E. O. W. Handbook of Flower Pollination

v. 1, p. 64. Oxford, 1906.

* The structure and development of the mango flower have been briefly discussed by Burns and Prayag. (Burns, W., and Prayag, S. H. Notes on the inflorescence and flowers of the mango tree. In Poona Agr. Col. Mag., March, 1911.)

ovary 1 millimeter broad, with a slender lateral style about 2 millimeters high. To one side and inserted upon the disk is the single fertile stamen, composed of a slender subulate filament about 1.5 millimeters long, surmounted by an oval, purplish red anther 0.5 millimeter long, which dehisces longitudinally. (Pl. II.) Occasionally two such stamens are produced. The whorl is completed by staminodes of varying prominence, short and subulate in some varieties, larger and capitate in others, some even becoming fertile and producing a few pollen grains. a few pollen grains. In the staminate flower the ovary is wanting.

ANTHESIS AND LIFE OF THE FLOWER.

Anthesis may take place at any time of the day or night, but it seems to be most frequent early in the morning. Usually a large number of buds will be seen to be opening about 6 or 7 o'clock and if the weather is clear the petals will be fully expanded and the anthers dehisced by noon. On exceptionally warm, bright days the anthers sometimes dehisce before the petals are fully expanded. Other flowers may open later in the day, and a few during the night.

The stigma has every appearance of being in a receptive condition immediately after anthesis, and in favorable weather retains its fresh appearance for about two days. At the end of this time the pistil usually commences to turn brown toward the stigmatic end, and the Ovary may commence to swell and assume a darker green color. In many flowers, however, the ovary swells but little, the flower drying up and falling off on the third or fourth day.

In flowers which have opened before noon, the pollen usually remains clinging to the anther for several hours, gradually being brushed off by insects or eventually falling. Sometimes much pollen will be left upon the anther until the second day. If the flower is staminate, usually it will shrivel and fall by the third day. Conditions seem to be vastly more favorable for pollination immediately after anthesis than at any later time, because of the larger amount of pollen present. No indications of heteracmy were observed, the stigma appearing to be receptive as soon as the flower opens, usually not more than an hour or two before the anther dehisces. It remains receptive for some time.

POLLINATION.

In spite of the close proximity of anther and stigma, the transfer of pollen from the former to the latter does not seem to be accomplished easily. Both the stamen and the pistil retain an erect position throughout, and the pollen as it is shed usually falls upon the base of the ovary or upon the disk rather than upon the stigma.

The normal method of transferring the pollen from the anther to the stigma must be through the agency of insects. The white,

INFLORESCENCE OF THE BRINDABANI MANGO.

This Indian variety produces larger panicles than many others, the length of the one here shown being about 14 inches. (Photographed by Wilson Popenoe at Miami, Fla., April 7, 1916; P16706FS.)

FLOWERS OF THE WHITE ALFONSO MANGO. SLIGHTLY ENLARGED.

Apex of a panicle, showing the character of the individual flowers, some which show the ovary, style, and stamen indistinctly. In this variety the perfect flowers are the first to open (at least toward the apex of the panicle), and are followed later by many staminate ones. (Photographed by Wilson Popenoe at Miami, Fla., April 7, 1916; P16703FS.)