A wise answer can only be given after a study of each case, and no one can do this so well as the planter himself. He should know, as no one else can know, his own conditions and requirements, and should be able to form very exact ideas of just what he wants, and the doing so is, in my opinion, one of the most important requisites for satisfactory tomato growing. I also believe that it is as impossible for a man to answer offhand the question, "What is the best variety of tomato?" as for a wise physician to answer the question, "What is the best medicine?"
Varietal names and descriptions mean something quite different in the case of plants like the tomato, which are propagated by seed, from what they do with plants like the apple and strawberry, which are propagated by division. In the latter case all the plants of the variety are but parts of the primal origination, and so are alike. A description is simply a more or less complete and accurate definition of what a certain immutable thing really is, but in the case of plants propagated by seed the variety is made up of all the plants which accord with a certain ideal. Bailey says, "Of all those which have more points of resemblance than of difference," and a description of the variety is of that ideal which in common practice is not fixed, but may and generally does vary not only with different people but from time to time. The only foundation for varietal names in plants of this class is an agreement as to the ideal the name shall stand for. Under modern horticultural practice when anyone has been able to secure seed most of which he is reasonably sure will develop into plants of a distinct type different from that of any sort known to him, he has a distinct variety, so that it is not surprising that we should find that American seedsmen offer tomato seed under more than 300 different names, and those of Europe under more than 200 additional, so that we have more than 500 varietal names, each claiming to stand for a distinct sort. Now it is quite possible—indeed, it is certain—that we might have 500 tomato plants each different in some respect, either of vine, leaf, habit of growth, or character of fruit, from any of the others and that these differences might make plants of one type better suited to certain conditions and uses than any other; but it is very certain that these 500 names do not stand for such differences. It is doubtless true that a portion—though I think but a small portion—of these different sorts exist simply as a matter of commercial expediency; but by far a greater part of them exist because one has found that plants of a certain character were better suited to some set of conditions and requirements than any sort with which he was acquainted, and having secured seed which he thought would produce plants of that character, has offered it as of a distinct sort.
It is probable that a better acquaintance with sorts already in cultivation would have prevented the naming of many of these stocks as distinct varieties. What is of far more practical importance, the same name does not always stand for precisely the same type with different seedsmen, or even with the same seedsmen in different years; nor are the seedsmen's published descriptions such as would enable any one to learn from them just what type he will receive under any particular name, or which sort he should buy in order to get plants of any desired type. Seedsmen's catalogs are published and distributed gratuitously at great expense, and are issued, primarily, for the sake of selling the seeds they offer. They answer the purpose for which they are designed, in proportion as they secure orders for seeds. Will this be measured by the accuracy and completeness of their descriptions? I think that it needs but slight acquaintance with the actual results of advertising to answer in the negative, and whatever your answer may be, the answer given by the catalogs themselves is an emphatic no.
In a recent case I looked very carefully through the catalogs of 125 American seedsmen who listed a certain variety which is very markedly deficient in a certain desirable quality, and found that but 37 of the 125 mentioned the quality in connection with the variety at all and of these but 7 admitted the deficiency, while 30 told the opposite of the truth. Even if a complete, exact and reliable description of a variety was published by disinterested persons, one could not be sure of getting seed from seedsmen which would produce plants of that exact type, since there is no agreement or uniformity among them as to the exact type any varietal name shall stand for.
One way of getting seed of the exact type wanted is to do as the South Jersey growers did: go to work and breed up a stock which is uniformly of the type wanted; but this involves more painstaking care than many are willing to give, though I think not more than it would be most profitable for them to expend for the sake of getting seed just suited to their needs.
A second and easier way is to secure samples of the most promising sorts and from the most reliable sources and grow them on one's own farm; select the stock which seems best for him and buy enough of that exact stock for several years' planting, and in the meantime be looking for a still better one. Tomato seed stored in a cool, dry place will retain its vitality for from three to seven years.