The chance finding of a single adult specimen of the solenogastre Neomenia carinata Tullberg 1875 rendered possible an embryological study of this species. Little is known concerning the ontogeny of the Aplacophora and a number of important questions, such as the fate of the larval test, the nature of the abapical depression visible during gastrulation and the presence or absence of any evidence of metamery, remain to be elucidated. Embryos, larvae and post-larvae were maintained in laboratory culture at 10 degrees C. No description can be given of the early cleavage stages since the eggs when found were always well advanced. Each egg is enclosed by a single membrane. Gastrulation begins on the second day, by a process of immigration of the abapical cells; the abapical depression, often called a blastopore, is shown to be of an unusual character and is to be referred to as a pseudo-balstopore. This pseudo-blastopore is merely a relatively shallow depression marking the area at which immigration is occurring. After the completion of gastrulation, the cells lining the pseudo-blastopore are the prospective trunk ectoderm. The endoderm and mesoderm lie within the embryo and have no communication with the exterior. The remaining cells form the larval test, except for an apical quartet which will develop into the larval apical plate and for six small patches of cells which will give rise to much of the definitive nervous system. The apical/abapical axis of the gastrula is coincident with the antero-posterior axis of the adult solenogastre. The embryos leave their egg membranes on the third day and swim by means of the cilia of the larval test. This test becomes organized into a series of tiers of regularly shaped cells. The main tier is the prototroch, on which is developed a strong equatorial band of locomotory cilia. The larvae are not negatively geotactic and swim close to the bottom of a culture vessel. Proliferation of the definitive nervous tissue begins just before hatching, from six areas of larval test ectoderm on the future ventral side. Nervous elements are cut off inwards at the bases of shallow ectodermal depressions; they come to aggregate into cerebral and ventral (pedal) ganglia. By the seventh day the rudiment of the adult trunk is visible, protruding through the pseudo-blastopore. On its tip is the yolk-laden, ciliated, larval telotroch. The remainder of the trunk is unciliated (except for a median longitudinal ventral ciliated band) but bears numerous pointed calcareous spicules. The length of the trunk rudiment increases by repeated division of the ectodermal cells within the pseudo-blastopore. The midgut passes down into the trunk and with it travel mesodermal elements and a pair of bands of nervous tissue which will form the ventral (pedal) cords. Proctodaeal and stomodaeal invaginations place the midgut in communication with the exterior but the larvae do not feed. The 'pygidial' development of the trunk of Neomenia resembles strongly that process as found in many annelids but it must be noted that no trace of metameric segmentation of this trunk is visible at any stage in the development of Neomenia. At no stage does the trunk bear overlapping dorsal spicules like those described by Pruvot for Nematomenia; it seems probable on embryological grounds that the solenogastres are more closely allied to the primitive Lamellibranchia than to the Polyplacophora. Metamorphosis is considered to include only those changes occurring from the tenth to the thirteenth days, during which period the larva exchanges a pelagic for a benthic life. The trunk comes to form by far the greater proportion of the late larva and swimming becomes impossible. The larval test cells lose their orderly arrangement, the prototroch ceases to exist as a co-ordinated locomotory organ and the whole larval test becomes enclosed within the blastocoel of the trunk by the anterior extension and fusion of folds of definitive ectoderm. Similarly, the larval telotroch enters the trunk blastocoel posteriorly. From the blastocoel these yolk-laden cells of the larval locomotory and sensory apparatus pass through the midgut wall into the digestive cells; here they are broken down into small clusters of yolk granules which form the main identifiable food reserve of the post-larva. The mouth and anus, which, before metamorphosis, were directed posteriorly, are now directed ventrally; they lie at the anterior and posterior extremities of a median ventral longitudinal ciliated groove, the so-called pedal groove of the post-larva. This groove is at no stage employed as a pedal sole. The sites from which nervous elements were proliferated during larval life are obliterated at metamorphosis. In the post-larva, two new pairs of ectodermal nervous depressions develop. Both give rise to tubular strands of nervous tissue which extend to and fuse with the cerebral ganglia. Lateral (pleural) cords develop as outgrowths from the cerebral ganglia. Post-larval stages lived in the laboratory without food for up to 10 weeks; they were subsisting entirely on their food reserves. The natural diet of the species is unknown. During the ninth week after metamorphosis the atrium appeared, as a capacious invagination around the mouth. No radula, gills, cloaca, heart, coelomoducts or gonads developed before the young stages died; all but the first are known to be present in the adult Neomenia. A bibliography of works dealing with the ontogeny of Aplacophora is given.