basis of relative precipitation during the preceding winters. Such a theory is supported by Font's statement (F2, p. 418) regarding the marshes and lagoons in Gilroy Valley: "Since it had not rained much this year it was quite dry...."

Examination of the modern maps is instructive. The U. S. Geological Survey quadrangles for the Bay Area were mapped in the decade 1895 to 1905, with subsequent re-editing. Thus the picture presented antedates the disruption of natural conditions caused by the population expansion of this century. The USGS sheets now show 20 streams between Milpitas and Rodeo which were named on the maps, or whose names are otherwise well recognized. These are, from south to north: Arroyo Coches Creek, Calera Creek, Scott Creek, Toroges Creek, Agua Fria Creek, Agua Caliente Creek, Mission Creek, Alameda Creek, San Lorenzo Creek, San Leandro Creek, the creek flowing past Mills College, the creek flowing out of Diamond Canyon through Fruitvale, Indian Gulch, Claremont Creek, Strawberry Creek, Codornices Creek, Wildcat Creek, San Pablo Creek, Pinole Creek, and Rodeo Creek. In addition there are approximately 12 unnamed creeks or arroyos descending the front slope of the hills onto the plain. The total, 32, conforms very closely with the descriptions left by the Spaniards in 1772 and 1776.

It is further to be noted that the government surveys of the period of 1900 showed all these streams as intermittent and therefore dependent for their flow upon the winter rainfall. Thus it is quite probable that in late March the amount of water might be copious (as in 1772) or relatively scanty (as in 1776). The conclusion seems warranted that, apart from dessication induced by such factors as soil damage, overgrazing, and diversion for minor irrigation projects, the local water supplies have not diminished since 1770. It then follows that under primitive conditions the natives had substantially the same quantity of water at their disposal as is available in the same area today.

3. The distribution of Indian population was quite clearly marked. Crespi (see above) encountered 5 villages from Milpitas to San Lorenzo. Anza mentions 6 villages in the same distance, most of which were north of Irvington. Both Crespi and Anza describe seeing no further villages until they reached Wildcat Creek. Thereafter they saw a village on each of San Pablo, Pinole, and Rodeo Creeks. Crespi puts 5 villages between Pinole and Crockett. Anza refers to only one, in addition to the one on Rodeo Creek, viz., the large village at Tormey. All accounts agree that there were no settlements between Crockett and Concord Valley, where 2 villages were found by both parties. On the return journey, Crespi traversed the valley from Walnut Creek to Dublin, Pleasanton, and near Niles, and noted a scattering of rancherias at least as far as Pleasanton. The Anza expedition, after leaving the delta, crossed the hills back of Mt. Hamilton and emerged near Gilroy, all without noting a permanent habitation site. From these accounts it is clear that the heavy concentration of population was along the Bay shore, locally centering on the large arroyos and avoiding the strip where Oakland and Berkeley now stand. Secondary centers were in the broader and lower interior valleys, west and north of Mt. Diablo. The narrow canyons and the brush-covered belts of the main axis of the Coast Range were destitute of inhabitants.

4. The numerical value of the population was estimated from the Crespi diary as 2,400. We may assay a comparable estimate from Font and Anza.

The six villages seen on the way from Warm Springs to San Lorenzo were largely abandoned by the inhabitants, who had fled in terror. Hence no population estimate is given. However an average of 100 persons each would be a reasonable assumption.

North of Berkeley an abandoned village was seen, probably on Wildcat Creek (possibly Strawberry Creek). Then a village of 100 persons was found on San Pablo Creek and another "fair sized" village probably on Pinole Creek. At Rodeo Creek was a village larger than the other two. Allowing 100 persons each for the abandoned village and the "fair sized" village and 150 for the one on Rodeo Creek, the four sites may be considered to represent at least 450 inhabitants. The village at Tormey had a population of 400, according to Font, and 500, according to Anza. No others are mentioned except the two in Concord Valley to which may be ascribed 100 persons each. If we use Font's estimate of 400 for the largest town, the total from Warm Springs to Concord Valley is 1,650. This is reasonably close to the value of 1,900 derived from Crespi.

For the remainder of the Northern Costanoan territory, since Anza found no inhabitants south of the Livermore Valley, we have to use the figures derived from Crespi: approximately 500 for the interior valleys. The total, then is 2,150 as compared with the 2,400 based upon Crespi's account alone. In any case, the present estimate is purely tentative and must be considered in the light of the Mission baptism records which are subsequently described.

THE CAÑIZARES EXPEDITION

During the period of initial land exploration attempts were made to secure information by water. The most important such episode was the voyage, if one may call it that, made by José Cañizares in 1775. Cañizares was the first mate of the ship, San Carlos, under command of Juan Manuel Ayala. Ayala was commissioned to survey the entire San Francisco Bay area, but was unable to complete the task because of illness. He therefore delegated the interior exploration to Cañizares, who fulfilled the mission in late August and early September. The results of the trip are embodied, first, in a series of maps, and second, in a letter by Cañizares (1775) to Ayala. (For historical background, the work of Cutter, 1950, may be consulted.)

The maps are three in number, all versions of the same map drawn by Cañizares, and dated respectively 1775, 1776, and 1781. The first is very poorly executed and shows little more than the outline of the Bay; it is not reproduced here.

The second (map 2) is carefully done and gives an extensive list of localities. The original is in the Ministry of War, Madrid, Spain. It is an elaboration of, and a very great improvement on, a map drawn by Ayala in 1775 which showed merely the outline of San Francisco, without detail. Ayala's map has slight value, hence it is not shown here, but a copy is available in the Bancroft Library, Berkeley.

Since the present map is itself reproduced from a photograph of a photograph, the text of the legends in the boxes is very faint and blurred. To facilitate reading, these legends have been copied, with translations. The symbols used on the map are the Latin alphabet, using capitals, for twenty-three items; they then continue as Greek letters, which are difficult to decipher and do not run strictly in sequence. Hence, for convenience, I have substituted in the legends numbers for the Greek letters, number 1 following Z of the alphabetical series. The use of these numbers in conjunction with the Greek letters on the map will not be difficult. In the left-hand column is the Spanish text; in the right-hand column, a literal translation. No attempt is made to correlate the names given by Cañizares with those applied at the present time.

The third map (map 3) copies the second, is carefully done, and gives essentially the same information, but varies in minor points. Cañizares remained in San Blas for several years after his visit to San Francisco. During this period his 1776 map was apparently redrawn by Manuel Villavicencio, in 1781, presumably under the supervision of Cañizares himself. Whether it is more accurate than the 1776 map is open to question. Small and capital letters are used for the localities on the map and in the legends.

The letter describing the survey of the Bay was written by Cañizares and addressed to "Señor Capitan," obviously