The larvæ, which become the minute parasitic males, undergo a very different and far less complete metamorphosis than those which become females. They attach themselves to the proboscis of an adult female, and lose their ciliated bands. Germinal cells make their appearance in the mesoblast, which form spherical masses, and, like the germinal balls in the female ovary, consist of a central cell, and an epithelium around it. The central cell becomes very large, while the peripheral cells give rise to the spermatozoa. A body cavity becomes developed in the larvæ, into which the spermatic balls are dehisced. Neither mouth nor anus is formed. The further changes have not been followed out. The larval males make their way into the oesophagus of the female, where they no doubt live for some time, and probably become mature, though the seminal pouch of the adult is not found in many of the males living in the oesophagus. When mature the males leave the oesophagus, and pass into the uterus. Phascolosoma. Cilia appear in Phascolosoma (Selenka, No. 369) while the ovum is still segmenting. After segmentation they form a definite band immediately _behind_ the mouth, which divides the larva into two hemispheres--a præ-oral and a post-oral. A præ-oral band of cilia is soon formed close to the post-oral band, and at the apex of the præ-oral lobe a tuft of cilia also appears. The larva has now the characters of a trochosphere, but differs from the typical trochosphere in the post-oral part of the ciliated equatorial ring being more important than the præ-oral, and in the absence of an anus. The metamorphosis commences very early. The trunk rapidly elongates, and the præ-oral lobe becomes relatively less and less conspicuous. The zona radiata becomes the larval cuticle. Three pairs of bristles are formed on the trunk, of which the posterior pair appears first, then the anterior, and finally the middle pair: an order of succession which clearly proves they can have no connection with a true segmentation. The tentacles become developed _between_ the two parts of the ciliated ring, and finally the præ-oral lobe, unlike what takes place in the Gephyrea armata, nearly completely vanishes. The anus appears fairly late on the dorsal surface, and the ventral nerve cord is established as an unganglionated thickening of the ventral epiblast. GEPHYREA TUBICOLA. The larva of Phoronis was known as Actinotrocha long before its connection with Phoronis was established by Kowalevsky (No. 372). There is a complete segmentation leading to the formation of a blastosphere, which is followed by an invagination, the opening of which is said by Kowalevsky to remain as the mouth[151]. It is at first terminal, but on the development of a large præ-oral lobe it assumes a ventral position. The anus is formed at a later period at the posterior end of the body.