While it is unnecessary to take soil samples oftener than once a day, it is important to make at least one series at each visit. Sometimes it becomes desirable to know the rate of water loss in different stations during the day, and in this event, samples are taken at one or two hour intervals for the entire day. In making simultaneous readings at the different levels of one station, the observers are grouped in one spot in such a way that they do not interfere with the correct reading of each instrument. Readings of this sort are most valuable in the case of temperature, which shows greater differences at the various levels. Important differences of humidity and wind also are readily obtained, and, in layered formations, marked variations in the amount of light. In the open, the ordinary levels for temperature are 6 feet, 3 feet, surface, 5, 10, and 15 inches in the ground, and for wind and humidity, 6 feet, 3 feet, and surface. In forests the same levels are used for comparison with formations in the open, but a more desirable series for light especially is secured by making readings at the height of, or better, just below the various layers. Series of this sort are likewise made on signal. The best time of day is that of a period in which the middle station is read near meridian, since the variation due to time is sufficiently small to permit fairly accurate comparisons between the readings for the different stations. =135. Records.= The form used for recording the observations made by means of simple instruments is shown below. It is hardly necessary to state that it may be readily modified to suit the needs of different investigators. Ordinarily, each sheet is used for the records of one habitat or series alone, but for convenience sake, the records of two different series are here combined. The figures given are taken from records for the prairie and forest formations at Lincoln. _Koelera-Andropogon-psilium_ April 25, 1901. Clear. South wind.