tionships; they may be 1) on the surface as "free sensory hillocks," usually with protruding cupulae (see review by Lowenstein, 1957); 2) sunk in pits below the skin, usually with protruding cupulae (the pit organs of Sato, 1955); 3) within elongated open furrows or enclosed canals, with cupulae not readily observed (Goodrich, 1930). Herrick (1899) studied in detail the lateral line system of an adult and larval Menidia, but he did not mention cupulae nor did Kuntz and Radcliffe (1916) who described the embryological and larval development of Menidia. Many of the groups, in addition to the Atherinidae of the order Mulgiliformes, in which cupulae have been found are summarized below. The fact that they are found in so many diverse groups suggests that they may be present in many fishes (Table 1), but have usually escaped notice in many because of their delicate structure, transparency, poor resolution properties, their location, frequently, within enclosed canals, and failure to be preserved. There is general agreement that cupulae are gelatinous, transparent structures which jut out from the epidermis. The earliest report of the lateral line cupulae describes them as "hyaline tubes" (Schulze, 1861); this tubular property was observed by Clapp (1898). Dijkgraaf (1934) argued that cupulae were not hollow, but were solid, jellylike, water-rich masses that enclosed sensory hairs (Dijkgraaf, 1952). Cylindrical cupulae were described for live Phoxinus (Dijkgraaf, 1934) and for live Fundulus (Denny, 1936 and 1937). Needle-shaped cupulae were described in several planktonic larvae by Thomopolous (1957).