We present the first measurement of the planet frequency beyond the "snow line," for the planet-to-star mass-ratio interval −4.5 < log q < −2, corresponding to the range of ice giants to gas giants. We find d 2 N pl d log q d log s = (0.36 ± 0.15) dex −2 at the mean mass ratio q = 5 × 10 −4 with no discernible deviation from a flat (Öpik's law) distribution in logprojected separation s. The determination is based on a sample of six planets detected from intensive follow-up observations of high-magnification (A > 200) microlensing events during 2005-2008. The sampled host stars have a typical mass M host ∼ 0.5 M , and detection is sensitive to planets over a range of planet-star-projected separations (s −1 max R E , s max R E), where R E ∼ 3.5 AU (M host /M) 1/2 is the Einstein radius and s max ∼ (q/10 −4.3) 1/3. This corresponds to deprojected separations roughly three times the "snow line." We show that the observations of these events have the properties of a "controlled experiment," which is what permits measurement of absolute planet frequency. High-magnification events are rare, but the survey-plus-follow-up high-magnification channel is very efficient: half of all high-mag events were successfully monitored and half of these yielded planet detections. The extremely high sensitivity of high-mag events leads to a policy of monitoring them as intensively as possible, independent of whether they show evidence of planets. This is what allows us to construct an unbiased sample. The planet frequency derived from microlensing is a factor 8 larger than the one derived from Doppler studies at factor ∼25 smaller star-planet separations (i.e., periods 2-2000 days). However, this difference is basically consistent with the gradient derived from Doppler studies (when extrapolated well beyond the separations from which it is measured). This suggests a universal separation distribution across 2 dex in planet-star separation, 2 dex in mass ratio, and 0.3 dex in host mass. Finally, if all planetary systems were "analogs" of the solar system, our sample would have yielded 18.2 planets (11.4 "Jupiters," 6.4 "Saturns," 0.3 "Uranuses," 0.2 "Neptunes") including 6.1 systems with two or more planet detections. This compares to six planets including one twoplanet system in the actual sample, implying a first estimate of 1/6 for the frequency of solar-like systems.

We analyze OGLE-2003-BLG-262, a relatively short, t E = 12.5 ± 0.1 day, microlensing event generated by a point-mass lens transiting the face of a K giant source in the Galactic bulge. We use the resulting finitesource effects to measure the angular Einstein radius, θ E = 195 ± 17 µas, and so constrain the lens mass to the full-width half-maximum interval 0.08 < M/M ⊙ < 0.54. The lens-source relative proper motion is µ rel = 27 ± 2 kms −1 kpc −1 . Both values are typical of what is expected for lenses detected toward the bulge. Despite the short duration of the event, we detect marginal evidence for a "parallax asymmetry", but argue that this is more likely to be induced by acceleration of the source, a binary lens, or possibly by statistical fluctuations. Although OGLE-2003-BLG-262 is only the second published event to date in which the lens transits the source, such events will become more common with the new OGLE-III survey in place. We therefore give a detailed account of the analysis of this event to facilitate the study of future events of this type.

Robust fast methods to classify variable light curves in large sky surveys are becoming increasingly important. While it is relatively straightforward to identify common periodic stars and particular transient events (supernovae, novae, microlensing events), there is no equivalent for non-periodic continuously varying sources (quasars, aperiodic stellar variability). In this paper we present a fast method for modeling and classifying such sources. We demonstrate the method using ∼ 86, 000 variable sources from the OGLE-II survey of the LMC and ∼ 2, 700 mid-IR selected quasar candidates from the OGLE-III survey of the LMC and SMC. We discuss the location of common variability classes in the parameter space of the model. In particular, we show that quasars occupy a distinct region of variability space, providing a simple quantitative approach to the variability selection of quasars.

In the favoured core-accretion model of formation of planetary systems, solid planetesimals accumulate to build up planetary cores, which then accrete nebular gas if they are sufficiently massive. Around M-dwarf stars (the most common stars in our Galaxy), this model favours the formation of Earth-mass (M % ) to Neptune-mass planets with orbital radii of 1 to 10 astronomical units (AU), which is consistent with the small number of gas giant planets known to orbit M-dwarf host stars 1-4 . More than 170 extrasolar planets have been discovered with a wide range of masses and orbital periods, but planets of Neptune's mass or less have not hitherto been detected at separations of more than 0.15 AU from normal stars. Here we report the discovery of a 5.5 15.5 22.7 M % planetary companion at a separation of 2.6 11.5 20.6 AU from a 0.22 10.21 20.11 M ( M-dwarf star, where M ( refers to a solar mass. (We propose to name it OGLE-2005-BLG-390Lb, indicating a planetary mass companion to the lens star of the microlensing event.) The mass is lower than that of GJ876d (ref. 5), although the error bars overlap. Our detection suggests that such cool, sub-Neptune-mass planets may be more common than gas giant planets, as predicted by the core accretion theory.Gravitational microlensing events can reveal extrasolar planets orbiting the foreground lens stars if the light curves are measured frequently enough to characterize planetary light curve deviations with features lasting a few hours 6-9 . Microlensing is most sensitive to planets in Earth-to-Jupiter-like orbits with semi-major axes in the range 1-5 AU. The sensitivity of the microlensing method to lowmass planets is restricted by the finite angular size of the source stars 10,11 , limiting detections to planets of a few M % for giant source stars, but allowing the detection of planets as small as 0.1M % for main-sequence source stars in the Galactic Bulge. The PLANET collaboration 12 maintains the high sampling rate required to detect low-mass planets while monitoring the most promising of the .500 microlensing events discovered annually by the OGLE collaboration, as well as events discovered by MOA. A decade of pioneering microlensing searches has resulted in the recent detections of two Jupiter-mass extrasolar planets 13,14 with orbital separations of a few AU by the combined observations of the OGLE, MOA, MicroFUN and PLANET collaborations. The absence of perturbations to stellar microlensing events can be used to constrain the presence of planetary lens companions. With large samples of events, upper LETTERS 1 PLANET/RoboNet Collaboration

Searches for extrasolar planets have uncovered an astonishing diversity of planetary systems, yet the frequency of solar system analogs remains unknown. The gravitational microlensing planet search method is potentially sensitive to multiple-planet systems containing analogs of all the solar system planets except Mercury. We report the detection of a multiple-planet system with microlensing. We identify two planets with masses of approximately 0.71 and approximately 0.27 times the mass of Jupiter and orbital separations of approximately 2.3 and approximately 4.6 astronomical units orbiting a primary star of mass approximately 0.50 solar mass at a distance of approximately 1.5 kiloparsecs. This system resembles a scaled version of our solar system in that the mass ratio, separation ratio, and equilibrium temperatures of the planets are similar to those of Jupiter and Saturn. These planets could not have been detected with other techniques; their discovery from only six confirmed microlensing planet detections suggests that solar system analogs may be common.