Most people count the "hits" and disregard the "misses." If such a record be kept, the wheat will be separated from the chaff, but a week or a month of record, while long enough to bring out much of interest, is far too short a time for purposes of scientific comparison of relative values.
Travelers, even when passing rapidly through a country on the railroad, and, still better, when moving more slowly on horseback or on foot, usually have opportunities for making simple non-instrumental observations which will add greatly to the interest of their journey, and which, if the region is comparatively little known, may really be of considerable importance. I have long felt that what I have termed "car-window climatology" deserves far more attention than it has received. In my own experience when traveling in South America under conditions which usually made it impossible to carry any instrument except a sling psychrometer, it was found feasible, when journeying on horse or mule back, on the Brazilian "trolley," or in the train, to collect facts which added greatly to my understanding of the climatology of the regions passed through; made the trips alive and interesting, and helped to hasten the passage of many weary hours. Some of these observations, indeed, it has seemed worth while to publish. There are many observations which an intelligent traveler can make, even from a car-window, although a slower method of progression is, of course, to be preferred in such a study. "Wind velocity may be reasonably accurately estimated, after a little practise, by noting the effect of the wind in blowing trees, or in producing waves of different sizes in lakes or on rivers. The prevailing wind direction can often be very accurately determined by observing the slant of wind-blown trees, or again, by taking note of the effects of wave action on the leeward side of a lake or pond. Vegetation always furnishes a general criterion in regard to temperature and rainfall. When trees shed their leaves we infer a season of cold or it may be of drought. The occurrence of frost may be detected both by seeing it, and by noting its effects. The altitude reached by frost may likewise be observed. The direction of rainy or snowy winds may be discovered by observing on which side trees are wet. Whether or not hail, or snow, or sleet, or frozen rain, or gales, or heavy rains, or fog, occur in a region is observable, so far as the period of his visit is concerned, by any traveler. Forest and prairie fires indicate droughts, or dry seasons. Tornadoes and high gales, may be detected many years after their occurrence by the damage they did to trees. Whether or not a river is subject to floods may usually be determined by such hurried observations as can be made from a car-window, by noting the mud deposited by former floods on the trunks of trees, or by seeing the banks and neighboring fields actually overflowed.