John O'Connell was a consultant neurosurgeon at St Bartholomew's. He was born in Manchester on 16 September 1906, to Irish parents. His father, Thomas Henry O'Connell, was a civil servant. His mother was Catherine Mary née O'Sullivan. He was educated by the Jesuits at Clongowes and later at Wimbledon College, before entering St Bartholomew's Hospital. There he won a junior scholarship in 1926, the Brackenbury surgical scholarship and the Wickes medal. After qualifying in 1931, he did junior posts at Bart's, was a demonstrator in anatomy and made a study of the anatomy of the cerebral veins and peripheral nerves, on which his later publications were to become the standard references. He became surgical chief assistant in 1937 under Geoffrey Keynes and Thomas Dunhill, an experience which convinced him of the value of extensive basic training in general surgery for future neurosurgeons, a view not shared by all his contemporaries. In 1935, a Rockefeller Foundation fellowship took him to the United States, where he worked at Ann Arbor with Max Peet and in Chicago with Percival Bailey, and during this visit was proud to have met Harvey Cushing, considered the father of neurosurgery. He suffered an attack of sciatica whilst in the States and learned of the work of Mixter and Barr on prolapsed lumbar intervertebral discs. On his return, he was one of the first to carry out their operation at Bart's, although he continued to be very conservative in its use. During the second world war he was in charge of the Emergency Medical Service neurosurgical unit at Hill End Hospital. Shortly before D-day he took his unit near to Portsmouth, where they treated more than 200 neurosurgical casualties. He was appointed to the staff of St Bartholomew's in 1946, but continued to work at Hill End until his new block could be built. There O'Connell became celebrated for the successful separation of three pairs of craniopagus Siamese twins in 1958, 1961 and 1964. He followed all his patients up indefinitely, so that his clinic was an interesting source of information about the long-term outlook for various conditions, especially acoustic nerve tumours, which he thought should only be removed when causing serious symptoms. He made a number of important and original contributions: on cordotomy for pain relief; a post-mortem study of the anatomy of the chiasm, explaining the loss of visual fields in pituitary tumours; and on the role of cervical spondylosis in cervical cord disease. He was twice an Hunterian Professor, President of his section of the Royal Society of Medicine and civilian consultant to the Royal Navy. Conservative in all things, O'Connell was President of the Society of British Neurological Surgeons at a time when formalised surgical training and the inspection of units was being mooted, to both of which he was steadfastly opposed. A quiet, private man with a strong Roman Catholic faith, John had a courteous and old-fashioned manner, was a keen follower of cricket and rugby, and a devoted fly-fisherman. He married his theatre sister, Marjorie Cook, in 1973. He died on 27 April 2001, aged 94.