SCIENTIFIC CORRESPONDENCE from around 2200 to 1200 BC, have been proposed for the Exodus tradition and emergence of Israel in Canaan, based on different interpretations of the archaeo- logical dataset (ref. 16). Nevertheless, although they are powerful tools, archae- ology and pottery are not the sole avenues that can be used to unravel the human past. Environmental events, high-preci- sion radiocarbon dating and precise regional dendrochronologies may provide information unobtainable through archae- ological associations 17 . New, high-preci- sion 14 C data series of XVIIIth Dynasty Egypt and Levantine Middle Bronze Age sites are required. Wiggle-matching link- ages with the Aegean dendrochronology 14 may lead to truly absolute dating. Hendrik J. Bruins Jacob Blaustein Institute for Desert Research, Social Studies Center, and Department of Geography and Environmental Development, Ben-Gurion University of the Negev, Sede Boker Campus, 84990 Israel Johannes van der Plicht University of Groningen, Centre for Isotope Research, Nijenborgh 4, 9747 AG Groningen, The Netherlands Marine gamete-release pheromones SIR - In contrast to the large body of experimental evidence for the existence of sex pheromones in marine inverte- brates, only a few chemical structures of these compounds have been elucidat- ed1·2. Here we report the isolation and identification of two gamete-release pheromones of marine invertebrates. By bioassay-guided fractionation and purifi- cation, we identified uric acid as the sperm-release pheromone of Platynereis dumerilii and inosine as the main compo- nent of the egg-release pheromone com- plex of Nereis succinea (Annelida, Polychaeta). These compounds contain the purine ring system, which has not as yet been identified in sex pheromones of any species. Within the marine invertebrates, the Nereidae are well suited for chemical investigation of sex pheromones because of their distinctive breeding behaviour. In both species, reproduction strictly depends on an exact spatial and temporal synchronization. High population densi- ties within restricted areas and the syn- chronization in time are responsible for the formation of swarms. The reproduction process occurs as fol- lows: (1) in these swarms mate recogni- tion is achieved by chemical signals 3 . (2) At the moment of recognition the male discharges the egg-release pheromone (ERP)4' 5 • (3) The female is stimulated to swim with high velocity in narrow circles surrounded by swarming males. After an 1. Kittredge, M., Terry, M. & Takahashi, F. T. Fish. Bull. 69, 337-343 (1971). 2. Zeeck, E., Hardege, J. D., Bartels-Hardege, H. & Wesselmann G. J. exp. Zoo/. 246, 285-292 (1988). 3. Zeeck, E., Hardege, J, Bartels-Hardege, H. Mar. Ecol. Prag. Ser. 67, 183-188 (1990). 4. Bailly-Marer, Y. Mar. Biol. 24, 167-169 (1974). 5. Lillie, F. R. & Just, E. Biol. Bull. 24, 147-160 (1913). 6. Townsend, G. Biol. Bull. 77, 306-307 (1939). 7. Boilly-Marer, Y. in Advances in fnvertebrate Reproduction Vol. 4 (eds Porche!, M., Andries. JC. & Dhainaut, A.) 494 (Elsevier, Amsterdam, 1986). 8. Jaenicke, L., Muller, D. G. & Moore, R. E. J. Am. chem. Soc. 96, 3324-3326 (1974). 9. Zeeck, E., Hardege ,J. D., Bartels-Hardege, H. & Wesselmann, G. Tetrahedron Lett. 31, 5613-5614 (1990). 10. Zeeck, E., Hardege ,J. 0., Bartels-Hardege, H., Willig, A. & Wesselmann, G. J. exp. Zoo/. 260, 93-98 (1991). 214 induction period of 10-40 seconds the female spawns. The discharged egg cloud contains the sperm-release pheromone (SRP). (4) Males achieve immediate fer- tilization by circling around the eggs and emitting sperm clouds. It is known that the coelomic fluid of the oppposite sex is the source of the gamete-release pheromones 6 • 7 . With ref- erence to the well-examined volatile sperm attractants in marine brown algae 8 , we first investigated volatile compounds of the coelomic fluid. These studies led to the elucidation of substances responsible for mate recognition in both species 2 · 3 • 9 • 10 • However, none of these compounds elicit- ed the gamete release, in contrast to the natural pheromone-containing coelomic fluid. Hence, in the present study we analysed the non-volatile, water-soluble compounds of the coelomic fluid for pheromonal activity. To follow the target substances during isolation and purification, we developed a reliable behavioural bioassay. We injected samples with a microlitre syringe just in front of a mature specimen swarming in a glass dish. The release of a visible sperm cloud or loss of the total amount of eggs represented a positive pheromone response. We tested chro- matographic fractions of the coelomic fluid both singly and in combination. (Details of methods are obtainable directly from the authors.) In the case of P. dumerilii , we identi- fied the bioactive single peak fraction finally obtained as uric acid. This SRP is effective at a threshold concentration of 0.6 μM (solubility of uric acid in sea water, 32%, pH 8.2, 23 °C: 180 μM). We measured the total amount of uric acid in the coelomic fluid of one female as 1.0 ± 0.1 μg. When we dissolved this quantity of uric acid in 2 cm 3 sea water (egg cloud volume) we found an actual pheromone concentration of 3.0 ± 0.3 μM. This five- fold excess is essential to ensure success- ful fertilization taking into account natural conditions such as the effects of dilution by turbulence and currents in sea water. In the case of N succinea, we identified inosin as the main component of the ERP- complex. (We are investigating the com- plete composition of the ERP complex separetely.) We are now trying to identify the SRP of N succinea and the ERP of P. dumerilii, so that we can compare males and females in the same species. Our preliminary results suggest that both uric acid and inosin are species- and sex-spe- cific. They represent the first identified gamete-release pheromones in marine invertebrates. Remarkably, the purine ring system has not as yet been reported to be a feature in sex pheromones. Erich Zeeck Institute for Chemistry and Biology of the Marine Environment, Tilman Harder Manfred Beckmann Carsten T. Muller Department of Chemistry, Carl van Ossietzky University of Oldenburg, 26111 Oldenburg, Germany A silicon sensor for 50 2 SIR - Initially synthesized about 30 years ago, porous silicon is now attracting con- siderable attention because it both photo- luminesces 1 and electroluminesces 2 in the visible portion of the optical spectrum. Although much of the recent interest in porous silicon is related to the possibility of generating an electroluminescent device having applications in the areas of display technology and information stor- age, we here suggest that the photolumi- nescent properties of this material make it attractive as a highly sensitive sensor for sulphur dioxide. Our suggestion is based on our recent observation that the photo- luminescence of oxide-coated porous sili- con is rapidly and reversibly quenched on a selective basis by gas-phase S02. It is well documented that exposing freshly etched porous silicon to various organic reagents results in non-selective quenching of the observed photolumines- cence3. We have also previously reported selective quenching by gas-phase and aque- ous Bronsted bases when a thin surface oxide is present 4 • The observed quenching is reversible; subsequent exposure to a Bronsted acid restores the photolumines- cence to its initial intensity. Partial quench- ing of the photosensitive luminescence by Bronsted bases results in a blueshifting of approximately 40 nm in the emission maxi- mum. As with the photoluminescence intensity, exposure to acid returned the emission maximum to its initial position. The base-quenching/acid-restoration cycle can be repeated many times. We now report that SO2 gas is another effective quencher of oxidized porous sili- NATURE · VOL 382 · 18 JULY 1996