To show that the response is indeed fromthe polymer within the gap, we studied the I-Vresponse as a function of photoexcitation witha Xe lamp (150 W). The I-V response for thepolymer-filled nanowire becomes slightlymore conductive upon Xe light exposure.During the backward scan, the device wasirradiated with the Xe lamp starting at –0.1 V(red arrows in Fig. 2A), and a change in slopein the I-V response was observed. The tran-sient conductance change between 1.1 nS inthe dark to 1.6 nS when irradiated is con-sistent with an increase in charge-carrier den-sity, which would be expected if the gap werefilled with the p-type polypyrrole (16).We report a novel lithographic process thatallows one to generate designed gap structureson nanowire templates. The process is remark-ably controllable, high-yielding, and easy toimplement. It does not require sophisticatedand expensive instrumentation and facilities,and it allows manipulation of an importantclass of structures that cannot be easily manip-ulated with conventional lithographic tools.Being able to make gap or notched structureswith nanowires with OWL and relatively in-expensive instrumentation will facilitate thestudy of the electronic properties of nano-materials and open avenues to the preparationof novel disk structures, which could be de-signed to have unusual optical properties as afunction of gap and metal segment size Ee.g.,plasmon waveguides (17)^.References and Notes1. B. D. Gates et al., Chem. Rev. 105, 1171 (2005).2. M. A. Reed, C. Zhou, C. J. Muller, T. P. Burgin, J. M.Tour, Science 278, 252 (1997).3. J. Reichert et al., Phys. Rev. Lett. 88, 176804 (2002).4. H. Park, A. K. L. Lim, A. P. Alivisatos, J. Park, P. L.McEuen, Appl. Phys. Lett. 75, 301 (1999).5. C. Z. Li, H. X. He, N. J. Tao, Appl. Phys. Lett. 77, 3995(2000).6. J. Xiang et al., Angew. Chem. Int. Ed. Engl. 44, 1265(2005).7. C. R. Martin, Science 266, 1961 (1994).8. D. Routkevitch, T. Bigioni, M. Moskovits, J. M. Xu, J.Phys. Chem. 100, 14037 (1996).9. S. R. Nicewarner-Pena et al., Science 294, 137 (2001).10. N. I. Kovtyukhova, T. E. Mallouk, Chem. Eur. J. 8,4354 (2002).11. A. K. Salem, M. Chen, J. Hayden, K. W. Leong, P. C.Searson, Nano Lett. 4, 1163 (2004).12. S. Park, J.-H. Lim, S.-W. Chung, C. A. Mirkin, Science303, 348 (2004).13. Materials and methods are available as supportingmaterial on Science Online.14. R. D. Piner, J. Zhu, F. Xu, S. Hong, C. A. Mirkin, Science283, 661 (1999).15. D. S. Ginger, H. Zhang, C. A. Mirkin, Angew. Chem.Int. Ed. Engl. 43, 30 (2004).16. S. Park, S.-W. Chung, C. A. Mirkin, J. Am. Chem. Soc.126, 11772 (2004).17. S. A. Maier et al., Nat. Mater. 2, 229 (2003).18. C.A.M. acknowledges the U.S. Air Force Office of Scientif-ic Research (AFOSR), Defense Advanced Research ProjectsAgency (DARPA), and NSF for support of this research.Supporting Online Materialwww.sciencemag.org/cgi/content/full/309/5731/113/DC1Materials and MethodsFigs. S1 to S423 March 2005; accepted 9 May 200510.1126/science.1112666Atlantic Ocean Forcing ofNorth American and EuropeanSummer ClimateRowan T. Sutton*and Daniel L. R. HodsonRecent extreme events such as the devastating 2003 European summer heatwave raise important questions about the possible causes of any underlyingtrends, or low-frequency variations, in regional climates. Here, we present newevidence that basin-scale changes in the Atlantic Ocean, probably related tothe thermohaline circulation, have been an important driver of multidecadalvariations in the summertime climate of both North America and western Eu-rope. Our findings advance understanding of past climate changes and alsohave implications for decadal climate predictions.Instrumental records show that during the 19thand 20th centuries, there were marked vari-ations on multidecadal time scales in the sum-mertime climate of both North America (1–4)and western Europe (5). In the continentalUnited States, there were significant variationsin rainfall and drought frequency (1–4), and ithas been suggested (1, 4) that changes in theAtlantic Ocean, associated with a pattern ofvariation known as the Atlantic MultidecadalOscillation (AMO) (6, 7), were responsible.If confirmed, such a link would be importantfor climate predictions because the AMO isthought to be driven by the ocean_s thermo-haline circulation (6) and may be predictable(8, 9). However, thus far the evidence for anAtlantic link is mainly circumstantial, beingderived from observations and showing corre-lation rather than causality. Clarifying whetherAMO-related changes in the Atlantic Oceanwere indeed responsible for the observed var-iations in North American summer climateand whether, in addition, there were impactson other regions is therefore an importantchallenge.Figure 1 shows the time series and patternof North Atlantic sea surface temperatures(SSTs) that characterize the AMO during theperiod 1871 to 2003 (10). There are AMOwarm phases in the late 19th century and from1931 to 1960; cool phases occur from 1905 to1925 and from 1965 to 1990. The spatial pat-tern shows anomalies of the same sign overthe whole North Atlantic, with the largestanomalies (s È 0.3-C) found just east ofNewfoundland.Fig. 1. (A)IndexoftheAMO, 1871 to 2003. Theindex was calculated byaveraging annual meanSST observations (29)over the region 0-Nto60-N, 75-Wto7.5-W.The resulting time se-ries was low-pass fil-teredwitha37-pointHenderson filter andthen detrended, also re-moving the long-termmean. The units on thevertical axis are -C. Thisindex explains 53% ofthevarianceinthedetrended unfiltered in-dex and is very similarto that shown in (1).(B) The spatial patternof SST variations asso-ciated with the AMOindex shown in (A).Shown are the regres-sion coefficients (-C per SD) obtained by regressing the detrended SST data on a normalized (unitvariance) version of the index.ABNatural Environment Research Council Centres forAtmospheric Science, Centre for Global AtmosphericModelling, Department of Meteorology, University ofReading, Post Office Box 243, Earley Gate, Reading RG66BB, UK.*To whom correspondence should be addressed.E-mail: [email protected] EPORTSwww.sciencemag.org SCIENCE VOL 309 1 JULY 2005115To identify the climate variations asso-ciated with the AMO, we considered a simplecomposite difference of observational sea-level pressure (SLP), precipitation, and surfaceair temperature (SAT) data between the warmphase from 1931 to 1960 and the subsequent30 years, 1961 to 1990, which were dominatedby a cool