Next to Sand Key, going east, is Middle Key, the smallest of the Dry Tortugas islands. It is little more than an oval patch of sand, about 80 feet long by 50 feet wide. Millspaugh states that Lansing found no vegetation on Middle Key whatever and that it is so low that, like Long Key, it is awash in rough weather. Notwithstanding the possi bility of the last statement, the author found several large tufts of dead Cakile on the key, its season having passed in July. It is reason able to suppose that the island has been built a little higher in the past 12 years by the current action, and especially so since the general map accompanying the paper shows that Middle Key is built, as are indeed the others, on quite extensive shoals scarcely submerged.
This is the most outlying key of the group since the disappearance of North and Northeast Keys a few years ago. It is the largest of the three keys stretching northeast of Garden Key and is about 4.5 miles from the latter key, about one-third of a mile in length and less than half that in width. It is almost entirely covered with vegetation and shows several associations as outlined in this paper. Cenchrus and Cakile are disseminated thickly along the shores, the latter particularly along the southeast shelf of dry sand-beach. Uniola and Chamcesyce
are in fairly thick growth on the northern and southern ends, with a sprinkling of Uniola all along the western side. Tournefortia, as large, well-grown bushes, occurs on both the east and west sides of the Wand, hut back Home distance from the shore. The northern half of the key is implanted rather thickly by two patches of Iva imbricate, separated by a band of Near the middle of the island is a single small group of Sem°la plants. Sesuvium occurs on the north and northwest shores. In a comparison with Lansing and Millspaugh's survey it is seen that the plants listed are exactly the same as those in the present survey; 8 species are given, and these are the same with the exception of Cakile fusiformis, which is perhaps the same systematic error noted before, C. lanceolate being the one noted on the island at present. Taking these facts as evidence, one may assume that in the 12 years gone by all the species have held their footing and some have increased in amount, but not at the expense of the others, and as this key is very rarely visited, the ecologic influences governing the distribution of the species have been undisturbed.