[Footnote 4: According to the opinion of a colleague of mine, a botanist, the results of the experiments of Phipson must be regarded as very doubtful, and some oxygen would appear to be indispensable for the growth of plants. We have to imagine the development somewhat as follows: As the earth separated from the solar nebula, its temperature was very high at first in its outer portions. At this temperature it was not able to retain the lighter gases, like hydrogen and helium, for a long period; the heavy gases, like nitrogen and oxygen, remained. The original excess of hydrogen and helium disappeared, therefore, before the crust of the earth had been formed, and the atmosphere of the earth immediately after the formation of the crust contained some oxygen, besides much nitrogen, carbonic acid, and water vapor. The main bulk of the actual atmospheric oxygen would therefore have been reduced from carbon dioxide by the intermediation of plants. The view that celestial bodies may lose part of their atmosphere is due to Johnstone Stoney. The atmospheric gases escape the more rapidly the lighter their molecules and the smaller the mass of the celestial bodies. On these lines we explain that the smaller celestial bodies like the moon and Mercury, have lost almost all their atmosphere, while the earth has only lost hydrogen and helium, which again have been retained by the sun.] This oxygen is an essential element for the production of animal life. As animal life stands above vegetable life, so animal life could only originate at a later stage than plant life. Plants require, in addition to suitable temperature, only carbonic acid and water, and these gases will probably be found in the atmospheres of all the planets as exhalations of their inner incandescent masses which are slowly cooling. The presence of water vapor has directly been established, by means of the spectroscope, in the atmospheres of other planets—Venus, Jupiter, and Saturn—and indirectly by the observation of a snow-cap on Mars. The spectroscope further gives us indication of the presence of other gases. There is an intense band in the red part of the spectra of Jupiter and Saturn, of wave-length 0.000618 mm. Other new constituents of unknown nature have been discerned in the spectra of Uranus and Neptune. On the other hand, there is hardly any, or at any rate only a quite insignificant, atmosphere on the moon and on Mercury. This is easily understood. The temperature on that side of Mercury which is turned away from the sun is near absolute zero. All the gases of the planetary atmosphere would collect and condense there. If, then, Mercury had originally an atmosphere, it must have lost it as it lost its own rotation, compelling it to turn always the same face towards the sun. Similar reasons may account for the absence of a lunar atmosphere.