It is obvious, however, that the great inequalities of the bottom of the sea; the existence of continents; the slopes of the coast, more or less steep; the different breadths of channels and straits; finally, the winds, the pelagic currents, and a crowd of local circumstances,--must materially modify the course of the tides. Nor is the moon the only celestial body which influences the rise and fall of the waters of the sea. We have already said that the sun asserts an influence on the waves. It is true that, in consequence of its great distance, this only amounts to a thirty-eight-hundredth part of that of the earth's satellite. The inequality which exists between the solar and lunar days--the latter exceeding the first by fifty-four minutes--has also the effect of adding to or subtracting from this force alternately. When the sun and moon are in _conjunction_ (Fig. 6), or in opposition, that is to say, placed upon the same right line, their attraction on the sea is combined, and a _spring_ tide is produced. This happens at the period of the _syzygies_--the period of new and full moon. At the period of the _quadrature_, or the first and last quarters, the solar action, being opposed to that of lunar attraction, tends to produce a sensibly weaker tide. [Illustration: Fig. 6. Lunar-Solar Tides.] These effects are never produced instantaneously; but, the impulse once given, it will continue to influence the tides for two or three days, the highest and lowest tides being nearly in the proportion of 138 to 63, or of 7 to 3. The highest tides occur at the _equinoxes_, when the moon is in perigee; the lowest at the _solstices_, when it is in apogee. In our ports, and along the coast, the water rises twice in twenty-four hours, when it is said to be high water; when it retires, it is low water: they are respectively the _flux_ and _reflux_ of the waves. The tide is retarded every day about fifty minutes, the lunar day being twenty-four hours fifty minutes of mean time. If, for instance, it is high water to-day at two o'clock in the morning, that of the next day will take place at fifty minutes past two. Low water does not occur, however, at the half of the intermediate time; the flux is more rapid than the reflux: thus at Havre, Boulogne, and at corresponding places on this side of the Channel, it takes two hours and eight minutes more in retiring; at Brest, the difference is only sixteen minutes more than the flux. The daily retardation of high water by the passage of the moon in the meridian, at the equinoxes, is a constant quantity for the same locality, which can be determined by direct observation.