As far as the Indo-European period is concerned it seems now to be agreed that there were then three seasons: for only the roots occurring in the words _hiems_, _ver_, and _summer_ recur in a greater number of the Indo-European languages. The much criticised statement of Tacitus about the Germans is therefore corroborated: “They know and name winter and spring and summer, but are ignorant of the name and the goods of harvest”[320]. Spring however is not equivalent to the other two seasons, for Indo-European antiquity certainly also divided the year into two parts, the cold and the warm seasons. The question whether the primitive Indo-European tribe had two or three seasons is therefore pointless, and that this is so will be readily understood by anyone who has become familiar with the overlapping and the instability of the seasons of the primitive peoples. The same phenomenon repeats itself in the addition of a fourth season. The Greeks complete the circle of the year with the three seasons winter, spring, and summer (χειμών, ἔαρ, θέρος), but in Homer the fruit-harvest, ὀπώρη, already appears with the pretensions of an independent season. Alkman has these four[321]. The principle of nomenclature is however different: the first three names are derived from climatic phenomena, ὀπώρα from the fruit-harvest. Now since four climatic periods are naturally to be distinguished--cold, warmth, and two transitional periods--the logical consequence is that the fourth season should also be referred to the climate, and indeed to the still unnamed period of transition between summer and winter. This period however does not coincide with ὀπώρα, but follows it. The latter term is therefore corrected to φθιν- or μετόπωρον; the ὀπώρα naturally persists as the fruit harvest, and Theophrastus[322] counts it in addition to the other four and thus gets five seasons. The same thing seems to have happened in the case of the Latin _autumnus_, although the process cannot be demonstrated. If the small seasons are included the circle may be still further extended. Thus the pseudo-Hippocratean treatise Περὶ ἑβδομάδων[323] gives seven seasons:--1, seed-time, σπορητός, from the early rising of the Pleiades to the winter solstice; 2, winter, until the late rising of Arcturus; 3, tree-planting, φυταλιά, up to the spring equinox; 4, spring; 5, summer, from the early rising of the Pleiades up to that of Sirius; 6, fruit-harvest, ὀπώρα, until the early rising of Arcturus; 7, autumn. This arrangement is certainly affected by the septenary system which pervades the treatise, but is founded on a popular basis: the smaller seasons, which otherwise pass into the greater, are given an independent position by the side of these. The system has not prevailed, it is true, but it affords a typical example of the instability of the seasons.