Embryogenesis is in some way a model system. It has always been distinguished bythe aeactitude, evn punctio, of its anatomical descriptions. An experiment by one of the great mastetrs ofembnyology could be made the tect ofa discourse on saentific method. But something is wrong, or has been wrong. There is no theory ofdeelopment in the sense in whichMendel-im is a theory thataccounts for the results ofbreeding experi-ments. There has therefore been littk sense ofprogresswn or timeliness about embryological research. Of many papm delivered at embryologica meetings, howeerf good they may be in themselves, one too oftenfeels that they mtnqhthave been delivered 5 years beforehandwithout making anyone much the wiser, or deferredfor 5 yeans without making anyone conscious ofagreat loss.-P. MEDAWAR