To determine this point we may either employ the determination of pore space by a comparison of the density of the soil constituents (see chap. 7, p. 107) with the volume weight of the soil; or we may measure directly the amount of water required to fill the pore-space. For the latter purpose it is only necessary to measure the amount of water (conveniently flowing from a graduated pipette) which, rising slowly from below in a U-shaped tube so as to expel all the air before it, is required to fill a definite weight or volume of the soil entirely full, so as to rise to its surface. We thus ascertain the amount of empty space existing within the soil,[73] which in the absence of water will ordinarily be filled by air. [73] Simple as this operation appears to be, it is found to be by no means easy to expel with certainty every small air bubble without resorting to means which would destroy the natural condition of the soil; such as boiling, or the use of the air-pump. These determinations cannot therefore lay claim to great accuracy. In most cultivated soils, as already stated, the air-space constitutes about 25% to 50% of their volume; and this space when filled with water represents what is commonly termed their _maximum water capacity_ or saturation point. It is of interest to know this, because it has been ascertained from experience that in order that plants may reach their best development, the capillary water present should not amount to more than 60%, or less than 40% of its maximum water-holding capacity; thus leaving about half the pore-space filled with air. This optimum, however, varies somewhat for different plants, some, like celery, being more tolerant of excess, and others being more tolerant of a deficiency of moisture, as is the _e. g._, egg-plant, originally a desert growth.