The reproduction of the horsetail.—It has been seen that some of the haulms do not bear branches, but are set aside for the production of the cones. These fertile haulms come up in March, before the green sterile haulms appear. When the cones are carefully examined they are seen to be covered by a number of shield-shaped scales. To the inner side of each scale (Fig. 154, 3, 4) are attached from five to ten boxes, each containing a large number of little grains or spores. The boxes burst open when they are ripe, about the end of March, liberating the spores; and the fertile haulms, having performed the one duty for which they were developed, at once die down. If the spores fall in a favourable situation they germinate, and each gives rise to a new plant. The new plant is, however, not a horsetail, but a small filmy prothallus, somewhat like the prothallus of a fern. Some of the prothalli produce male cells, while others give rise to female cells. When the minute male cells are ripe, they are set free, and are able, by means of fine, lashing threads, to swim towards the female cells of a neighbouring prothallus, through a drop of dew or rain. The two cells fuse together and give rise to a little embryo, which in due course grows up into a new plant—an ordinary horsetail with stems and branches and roots like its grandparent. [Pg 196] The advantage of an alternation of generations.—The life-history of the horsetail is evidently very similar to that of a fern, each exhibiting a well-marked alternation of generations. There is reason for believing that the prothallus-generation is the original form, and that the ordinary fern and horsetail were developed simply to scatter spores at intervals, and so to give new plants the advantages of fresh soil. Fig. 155.—Calamite: an extinct plant allied to the Horsetails (greatly reduced).