Notwithstanding suggestions to the contrary, the branchiopod carapace primitively originates as a fold that arises at the posterior margin of the head. This is generally from the maxillary segment but this region has sometimes been so modified during the radiation of the group that it is not always possible to be so precise. Fusion of any part of the carapace with the thorax is rare and does not negate the origin of the carapace fold. Although this has been denied, the Cambrian Rehbachiella displays the primitive pattern to perfection. In at least some notostracans there is a minor connection with the anterior part of the first thoracic segment which is interpreted as a secondary association. Notwithstanding differences in early ontogeny in the Spinicaudata and Laevicaudata, the carapace is clearly of maxillary origin and, although their ontogeny is again different, this is so also in the Anomopoda and Ctenopoda. In the Haplopoda the carapace lies topographically posteriorly on what appears to be a thorax made up of six fused segments. Nevertheless it is of anterior origin. To facilitate the efficient use of its grasping limbs and to enable the highly modified thoracic musculature to be anchored on the extremely thin cuticle, the thorax of Leptodora (the only extant haplopod) has elongated dorso-ventrally and rotated. This directs the limbs forward and has carried the carapace posteriorly, though it still lies in a morphologically anterior position. Developmental studies reveal that it originated from the maxillary segment. The most highly modified branchiopod carapaces are found in the Onychopoda. Here they form a specialized brood chamber, always provided with a nutrient-secreting Nahrboden (or `placenta'). Most specialized of all are those of the podonids where what Claus called an `uterus' holds the eggs and developing young. These specialized structures are also derived from a classical carapace fold. There is no such thing as a `secondary shield', a term that has been used instead of carapace in the Spinicaudata, Laevicaudata, Anomopoda, Ctenopoda, Haplopoda and Onychopoda. As no such character exists, to use its alleged common possession as a synapomorphy which unites all these disparate orders into one phyletic lineage has no meaning. The Anostraca lack a carapace, as did the Devonian Lipostraca, and there is nothing to indicate that they ever had one.