Indus–Mesopotamia relations are thought to have developed during the second half of 3rd millennium BCE, until they came to a halt with the extinction of the Indus valley civilization after around 1900 BCE. Mesopotamia had already been an intermediary in the trade of lapis lazuli between the Indian subcontinent and Egypt since at least about 3200 BCE, in the context of Egypt-Mesopotamia relations. Neolithic Revolution and Neolithic A first period of indirect contacts seems to have occurred as a consequence of the Neolithic Revolution and the diffusion of agriculture after 9000 BCE. The prehistoric agriculture of the Indian subcontinent is thought to have combined local resources, such as humped cattle, with agricultural resources from the Near East as a first step in the 8th–7th millennium BCE, to which were later added resources from Africa and East Asia from the 3rd millennium BCE. Mehrgarh is one of the earliest sites with evidence of farming and herding in the subcontinent. At Mehrgarh, around 7000 BCE, the full set of Near Eastern incipient agricultural products can be found: wheat, barley, as well as goats, sheep and cattle. The rectangular houses of Mehrgarh as well as the female figurines are essentially identical with those of the Near East. The Near-Eastern origin of South Asian agriculture is generally accepted, and it has been the "virtual archaeological dogma for decades". Gregory Possehl however argues for a more nuanced model, in which the early domestication of plant and animal species may have occurred in a wide area from the Mediterranean to the Indus, in which new technology and ideas circulated fast and were widely shared. Today, the main objection to this model lies in the fact that wild wheat has never been found in South Asia, suggesting that either wheat was first domesticated in the Near-East from well-known domestic wild species and then brought to South Asia, or that wild wheat existed in the past in South Asia but somehow became extinct without leaving a trace.