Silicon based nano-photonics is becoming a prominent contender in the race for effective all-optical information processing - and is simultaneously becoming a fascinating arena for fundamental research. Integration of optical devices into microelectronic chips is now not only discussed by academic researchers, but is also included in the business plans of microelectronics giants such as Intel and IBM. One of the first practical applications of on-chip nanophotonics is likely to be compact optical processing of multifrequency data streams. Nano-sized silicon waveguides (photonic wires) and resonators offer a very attractive way of realizing photonic components on a chip. This is due to the large index contrast between silicon and air, so that light at a wavelength of 1550nm can be tightly confined for waveguide widths as small as 500 nm. Another widely-recorgnised advantage is the possibility of using well established and wide-spread complementary metal-oxide-semiconductor (CMOS) technology. The presence of a substantial ultrafast Kerr nonlinearity in Silicon nano-structures potentially allows devices to perform at the THz rates that will be required in near-future high-performance sub-systems. Nonlinearity and dispersion control are the key properties needed to develop all-optical processing devices such as modulators, switches, delay lines and amplifiers. They are also the key parameters to be controlled if we are to understand and explore the fundamental optical physics in these structures. The interplay between dispersion and nonlinearity leads to such effects as soliton formation and modulation instability, which will be essential for temporal control and spectral modification. One of the dreams of the optical soliton community has been a three dimensional photonic chip made of a nonlinear material where all the routing is done by means of the spatial solitons, which then serve as an instantly reconfigurable and flexible network of waveguides for transmission and processing of data by means of temporal solitons. The soliton effects in planar silicon chips proposed here are possibly as close as we can hope to get to this dream. The overall aims of the research programme are: to fabricate a range of silicon-on-insulator structures for observation of spatiotemporal solitons, frequency conversion, and spectral, temporal and spatial shaping of femtosecond pulses;bistability effects in cavity arrays; experimentally observe and model the above effects, develop their physical understanding; use the unique properties of silicon to observe new optical phenomena; ensure further scientific progress in the area of nonlinear nano-photonics.