The avocados cultivated in the United States are classified horticulturally in three races: the West Indian, the Guatemalan, and the Mexican. The West Indian and Guatemalan races, so far as can be judged at present, are two expressions of one botanical species, Persea americana, while the Mexican race represents a distinct species, Persea drymifolia.

Horticultural varieties of the avocado, when propagated from seed, do not reproduce the parent fruit in every detail. Seedlings from a round green fruit of the West Indian race may produce fruits oblong or pyriform in shape, and red or purple in color, varying from the parent in numerous other ways as well. But these seedlings will always be like their parents in certain respects, because they belong to the same race and will reproduce the racial even though not the individual characteristics.

To use the definition of H. J. Webber,1 "Races are groups of cultivated plants that have well-marked differentiating characters, and propagate true to seed except for simple fluctuating variations." Technically speaking, the Mexican avocados should not be called a race, since they really represent a species; the West Indian and the Guatemalan, however, do not appear to differ from each other except in minor characters.

The classification of avocados into these three races has been useful, inasmuch as it brings together all those varieties which have several characteristics in common. In fact, the mere statement that an avocado belongs to the West Indian, Guatemalan, or Mexican race gives one an idea of the relative hardiness, season of ripening, and commercial character of the fruit. The botanical standing of the cultivated races, as at present understood, and the characters which serve to distinguish them horticulturally, are shown in the following key:

1 In the Standard Cyclopedia of Horticulture.

1. Leaves anise-scented; skin of fruit thin (rarely more than 1/32 inch in thickness) Persea drymifolia Mexican Race of horticulture

2. Leaves not anise-scented; skin of fruit thicker (from 1/32 to 1/4 inch in thickness) Persea americana a. Fruit summer and fall ripening; skin usually not more than 1/16 inch thick, leathery in texture. West Indian Race b. Fruit winter and spring ripening; skin 1/16 to 1/4 inch thick, woody in texture. Guatemalan Race

One variety cultivated in the United States, the Fuerte, appears to be a hybrid between the Mexican and Guatemalan races. Others of similar origin are likely to appear at any time, hence it is desirable to establish a group to include hybrids.

The avocados of the West Indian race have been developed in the tropical lowlands; the Guatemalan race, on the other hand, is a product of the highlands. At intermediate elevations varieties appear which belong to neither of these races, but possess some of the characters of each. These intermediate forms cannot be classified with accuracy.

In selecting varieties for commercial planting, it must be borne in mind, first of all, that the tree must be vigorous and hardy enough to grow successfully in the particular location which the planter has in view. Secondly, it must in time produce sufficiently large crops of marketable fruit to make its culture commercially profitable. It is not necessary that it be very precocious; it is noticeable, in fact, that precocious varieties often fail to make vigorous trees. It is more desirable to have the tree devote itself during the first three years to the development of an extensive root-system and a well-branched crown capable of withstanding the drain imposed by the production of heavy crops of fruit than to have its growth limited and its vitality exhausted by premature fruiting. Thirdly, the fruit itself must be given consideration from a commercial standpoint. Attractiveness, flavor, shipping qualities, season, and other important characteristics should be considered in respect to the market it is proposed to supply. Naturally, good shipping quality can be sacrificed to some other point if the fruit is for local use, while it is essential if the fruit is destined for distant markets. The flavor and quality of the flesh should be as good as possible, and the seed should not be unduly large.

More than one hundred and fifty varieties have been propagated in the United States up to the present time. The larger part of these originated as seedlings in California and Florida; the remainder have been introduced from Mexico, Guatemala, Cuba, the Bahamas, Hawaii, and a few other regions.

Of this large number not more than a dozen are likely to be planted ten years hence. Indeed, most of them have already been discarded. New varieties are originating every year, however, and the introduction of promising sorts from foreign countries is receiving much attention. It is only by testing a large number of varieties from all of the important avocado regions of the tropics that the best available kinds for commercial cultivation can be obtained.

It is not desirable to burden such a work as this with descriptions of all the avocados which have been propagated. It is sufficient to include the more important ones which are at the present time being planted commercially. For descriptions of minor varieties, and for information regarding the behavior and value of new introductions, the reader is referred to the annual reports of the California Avocado Association. In 1917 this organization issued Circular No. 1, "Avocado Varieties Recommended for Planting in California," the suggestions contained in which have done much to eliminate from consideration numerous inferior sorts. The varieties recommended in this circular are as follows, the arrangement being according to season of ripening in California:

Spring varieties

Fuerte, Spinks, Blakeman, and Lyon

Summer varieties

Spinks, Blakeman, Lyon, Dickinson, and Taft

Fall varieties

Taft, Dickinson, and Sharpless

Winter varieties Sharpless, Puebla, and Fuerte

Several of these varieties may be superseded within a short time by others which are now being tested in California. It is not to be expected that the industry can settle down to the cultivation of a few standard sorts until all of the promising ones have been tested, and this may require several years.

In Florida, the only variety which was extensively planted during the first fifteen years of the industry was Trapp. With the introduction of the Guatemalans, however, the question has become more complicated, and it will take some time to determine by actual trial which members of this race are most suitable for cultivation in different parts of the state.

It is probable that varieties will be obtained which will make it possible, both in California and Florida, to market avocados in every month of the year. Indeed, it is almost possible to do so at the present time. In other regions horticulturists should work toward this end by obtaining for trial varieties ripening at different seasons.

Plate IV. Left, Puebla avocado tree producing its first crop at two years of age; right, the Fuerte avocado.