together with the muscles which operate them, comprise the chief engine of digging. This engine operates in powerful fashion in cutting away the earth, so as to make possible the rapid extension of the gopher's underground system of passageways. The adequate housing of the heavy incisor teeth, and the need of meeting the severe stresses during the action of the muscles which operate the jaws, have resulted in the great thickening and ridging of the bones of the skull. We find that the forefeet are larger than the hind feet, a reversal of the ratio in animals which can run with agility; and the forefeet are provided with long stout curved claws. The forearm and shoulder are heavily muscled, and thus the actions of the jaws and teeth are supplemented, in loosening and particularly in transporting the soil. So far as is known, no pocket gopher goes into dormancy at any season; none either aestivates or hibernates. The source of food upon which the pocket gopher can depend, year in and year out, and which it can seek in safety, is comprised in the underground stems and root stalks of various grasses and herbs. These it gets almost altogether by digging its way to them; it gathers food only as it can advance under cover. While it is true that gophers do pull into the temporarily open mouths of burrows, stems and leaves of above-ground plants, these latter, I am led to believe, constitute only a minor fraction of the total annual food supply of the animal. The only dependable food source, continuing throughout the year, is comprised in underground stems and roots. And this is an exceedingly important consideration in our present study; for the general geographic limitation of Thomomys, to North America west of the 100th meridian, coincides with the territory where sharp alternation of dry and wet seasons is characteristic of the climate. Linked up with this climatic peculiarity there is undoubtedly, in the Southwest, relatively greater abundance of plants with nutritious roots and thickened underground stems, which tide over the dry season, than in the remainder of North America, where there is no long dry season. In other words, the ancestral pocket gophers of the remote past made the fortunate discovery of an oncoming type of food source correlated with the increasing aridity of what came to be a marked climatic and vegetational province. Restricting our attention now to Thomomys as the genus occurs in California, I will revert to the fact of its well-nigh universal distribution within the State. How can the fact of this wide distribution be harmonized with the restriction in the animal's mode of existence which we have just pointed out in some detail? Examination of the territory wherever pocket gophers thrive, from one end of the State to the other, does show most emphatically close concordance of occurrence with those very, and special, conditions-of suitable food, and of consistency of soil which permits of digging. In other words, these two critical factors are widespread, and wherever they extend pocket gophers have gone. The hindrances, locally, to the spread of pocket gophers are comprised in, not altitude, not cold, not heat, but in discontinuity of ground wherein the pocket gopher can extend its burrows; in discontinuity of ground in which sufficient food of the kind the pocket gopher can use is available throughout the year; and of course, in impassable bodies or streams of water. In other words, we find operating as outright barriers to their distribution only ground such as lava flows which cannot be penetrated by gophers, or ground which is too dry or too alkaline to support adequate plant growth for the gophers' food throughout the year, or permanent streams or bodies of water which the gopher cannot cross. In this latter connection, the pocket gopher can thrive, we know, without ever drinking; in many parts of the State the only water it can get for long periods is contained in the plant tissues which it uses for food. On the other hand, the animal can live healthily in soil that is saturated with water. Yet it is forced out of the ground when the land is flooded, as during very heavy rains or when under irrigation. Not only a river itself, but the adjacent bottomland subject to overflow at high water, may thus be effective in limiting the spread of gophers locally. A gopher can swim short distances when forced to; but it does not take to water voluntarily. These facts bear on the problem of geographic differentiation of races now to be discussed. I have pointed out that, despite the pocket gopher's extreme specializations in structure and habits, despite its restriction to a very narrow range of living conditions, yet the fact that these special living conditions are widespread has permitted of the very wide distribution in California, of this type of rodent. Now we come to deal with the observation that while our pocket gopher as a genus exists in every county of California, from below sea level to almost the highest altitudes, from the hottest to the coldest portions of the State, and from the driest to the wettest belts, yet the species represented under all these varying conditions is not the same; the genus is broken up, as indicated on the accompanying map, into not less than 33 different races (species or subspecies), no two of them occupying precisely the same territory. And this fact signifies that the varying combinations of conditions resulting from the topographic and climatic diversity in California have made their impress upon the gopher sub-stocks, which are more or less isolated from one another in what may be called differentiation provinces. It is our problem to enquire as to what factors among the present or recently past conditions have resulted in this isolation and consequent differentiation of all these various stocks. As will be observed from the map, the most conspicuous gopherless areas in California lie in the southeastern desert territory, chiefly on the Mohave Desert. Extended explorations of that arid territory have been made, with the special object of determining the kinds and numbers of rodents and other mammals present. Almost every square mile of those deserts, save on such evaporation floors as those of Searles "Lake" and Panamint and Death valleys (where there is a heavy deposit of saline substances, and no chance of plant growth), supports a large population of seed-gathering rodents-kangaroo rats, pocket mice, and ground squirrels |