My idea is to obtain separate collections from all the principal rivers of Spain and Portugal, and even to have several separate collections from the larger rivers, one from their lower course, one from their middle course, and another from their head-waters. Take, for instance, the Douro. One collection ought to be made at Oporto, and several higher up, among its various tributaries and in its upper course; say, one at Zamora and Valladolid, one at Salamanca from the Tormes River, one at Leon from the Esla River, one at Burgos and Palencia from the northern tributaries, one at Soria and Segovia from the southern tributaries. If this could be done on such a scale as I propose, it would in itself be a work worthy of the Spanish government, and most creditable to any man who should undertake it. The fact is that nothing of the kind has ever been done yet anywhere. A single collection from the Minho would be sufficient, say from Orense or Melgaco. From the northern rivers along the gulf of Biscay all that would be necessary would be one thoroughly complete collection from one of the little rivers that come down from the mountains of Asturias, say from Oviedo. The Ebro would require a more elaborate survey. From its upper course, one collection would be needed from Haro or Frias or Miranda; another from Saragossa, and one from its mouth, including the minnows common among the brackish waters near the mouth of large rivers. In addition to this, one or two of the tributaries of the Ebro, coming down from the Pyrenees, should be explored in the same manner; say one collection from Pampeluna, and one from Urgel, or any other place on the southern slope of the Pyrenees. A collection made at Barcelona from the river and the brackish marshes would be equally desirable; another from the river at Valencia, and, if possible, also from its head-waters at Ternel; another from the river Segura at Murcia, and somewhere in the mountains from its head-waters. Granada would afford particular interest as showing what its mountain streams feed. A collection from the Almeria River at Almeria, or from any of the small rivers of the southern coast of Spain, would do; and it would be the more interesting if another from the river Xenil could be obtained at or near Granada, to compare with the inhabitants of the waters upon the southern slope of the Sierra Nevada. Next would come the Guadalquivir, from which a collection should be made at San Lucar, with the brackish water species; another at Seville or Cordova, one among the head-waters from the Sierra Nevada, and another from the mountains of the Mancha. From the Guadiana a collection from Villa Real, with the brackish species; one from Badajoz, and one from the easternmost headwaters, and about where the river is lost under ground.