I will now proceed to explain the process of raising new varieties by cross-fertilization. Supposing, for the sake of illustration, it was desired to endeavour to get a large scarlet variety which would produce four bloom sprays in abundance - which no scarlet variety at present does - one would ask oneself which was the brightest scarlet. The answer might be Dobbie's Scarlet, Red Star, or Scarlet Monarch. Well one of these would be one of the parents, while the other would be a vigorous growing variety which produced large flowers on four bloom sprays every time. Asking oneself which varieties did this, irrespective of colour, one would answer, Mrs. Cuthbertson, Agricola, Hercules or Marks Tey. One of these, say Mrs. Cuthbertson, would be adopted as the female parent, and in the early bud stage, some flowers would require to be emasculated by removing all the anthers before they had dehisced, i.e., burst and shed their pollen. These flowers must be carefully marked and the next day, or the following one, pollen must be brought from the other parent, the bright scarlet one, and applied to the stigma of Mrs. Cuthbertson. How this is best done is shown in the accompanying illustration. The marked flowers should then follow their normal process and ripen pods of seed which must be carefully saved and kept in separate packets till sowing time. The seeds are sown and treated in the ordinary way, special strong culture not being advisable for such plants. They ought however to be planted where they can have room to grow as distinct plants and not get intertwined with any others. It is here that one of the most wonderful and interesting things happens. On the flowers produced on these plants no dependence at all can be placed. Students of Mendel's law will understand why this is, and those who have not studied Mendelism ought to do so if they seriously intend to take up the raising of new Sweet Peas. The seeds of the plants produced in this, the first or F 1 generation, must be carefully saved and carefully kept separate and labelled. Next sowing time they, or a part of them, must be sown and planted at least a foot apart - eighteen inches is better - in the lines. In this generation, called F 2, new forms will arise and it is possible the large scarlet, giving plenty of "fours," which was the object in view when setting out, may appear. If it does, it must be carefully marked and the seed saved for sowing again next season to see if it is fixed, and will breed true to the improved type. Of course it may or it may not. It if does not, some other selection may give something of promise, and the process of growing to prove must be carried on the following season. It will be seen from this that the process of raising is a prolonged one, and after the new variety has proved to be fixed, a stock must be worked up for sale which may take other two or three years. The process of raising new Sweet Peas is very much like that of raising new Daffodils. It is the first years that are the trying ones. After persistently making crosses for four or five years, if one continues making a few each year, thereafter the reward comes along every year, because new forms come into bloom for the first time every season. Another hint I may give. Do not throw away odd things - give them a trial, as they often are the carriers of real novelties which appear later. It is of course here that experience tells, because an expert can often see what is likely to come out of a peculiarly tinted flower, though at the moment of observation it gives little promise.
One of the most lovely Sweet Peas ever raised is Audrey Crier, a most charming salmon pink; but it is a heterozygote and will not breed true. Those who admire this flower could have it true every year if they would make a cross between Princess Victoria and Edrom Beauty. In the first generation the flowers would be Audrey Crier, but in the next it would break up and give the three forms, Princess Victoria, Edrom Beauty and Audrey Crier. If seed was saved from Audrey Crier in the second generation, in the following one it would break up and give the other two as well as itself, hence its being termed heterozygous. Edrom Beauty and Princess Victoria if saved would breed true and are consequently termed homozygous.