There is in this case a duplication which affects the inferior part of the embryo; the primordial pelvis, with its appendages, is duplicated.

From the observations of Cleland, it appears that the two halves present peculiar and interesting relations. The duplication is in the middle line, so that a right and a left pelvis result. As the two pelves are closely connected behind (by the spinal column), there is much more room for expansion in front, and so while the proximate posterior parts are crushed together and liable to imperfect development, the anterior halves attain full development. The result is, two perfect limbs, which, however, belong properly to two different pelves, being the right leg of the right pelvis and the left leg of the left; and in addition, placed posteriorly, a more or less imperfectly formed composite pelvis with supernumerary legs, variously coalesced or dwarfed.

The condition of the supernumerary parts behind varies considerably. There is usually one coalesced limb, like that mentioned above; but there may be two, or the indications of a second.

This subject has been worked out by Cleland in his memoir on "Birds with supernumerary legs," etc. He there refers to cases in the human subject, especially to that of Dos Santos, a monstrosity which has been described by several writers. In this case there was a supernumerary leg attached posteriorly. It had the knee turned backwards, and ended in a composite foot with the great toes joined and the fifth toes at the outer sides. This composite limb, therefore, had not arisen by the fission of two limbs of the same pelvis, but by the coalescence of the left leg of a right pelvis and the right leg of a left pelvis. In front there were two penes between the fully-developed limbs, another indication that the anterior parts and the two developed limbs belonged to different pelves.