At CERN, the neutrino community has recently reflected on a broad variety of options
for future oscillation experiments.
Several promising ideas, both for 
and 
oscillation searches,
have been presented and discussed in a series of meetings and workshops held
at CERN.
Some of these have been published [33, 34, 35].
In this document we demonstrate that we can design an experiment which has high sensitivity
for oscillation
and which will be capable of contributing significantly in clarifying the experimental scenario.
This experiment will be ready to extend the investigation started
with CHORUS and NOMAD, improving the sensitivity
both in the mixing angle and in 
.
A combination of different approaches presented in the workshops
has been retained as a baseline design.
It was pointed out [36]
that a segmented emulsion target with passive material could allow a
considerable increase of the target mass of the experiment.
A solution (TENOR) was presented [34]
to fit a segmented fully-sensitive large-mass ( 
5 tons) emulsion target in a magnetic field,
complemented by
high-precision silicon and emulsion trackers and by other more conventional
electronic detectors.
The use of silicon detectors on a large scale, as proposed in
[33],
allows high accuracy in the extrapolation of particle tracks to the emulsion target.
With a target mass in the order of a few tons, and exploiting the excellent
and still improving performance of
the SPS neutrino beam (see section 3), it is possible to collect
about 
charged-current neutrino interactions in three years.
This large number of events can be scanned in its entirety without the need of
kinematical preselection [34]
owing to the recent advances in the technique of automatic emulsion scanning.
This technique, pioneered by the group of Nagoya [37],
has recently undergone further improvements.
The Nagoya and Salerno groups of the CHORUS Collaboration have developed second generation
automatic scanning systems [38, 39]
able to cope with the request of the proposed experiment.
The high spatial resolution of the silicon detectors can be used to restrict the surface
area to be scanned, thus resulting in a significant reduction in scanning time [40].
A large rejection of background can be achieved by kinematical analysis of the
candidate events using the knowledge of the direction of the 
before its decay.
The need for a high efficiency on the measurement of momenta of charged tracks
makes it mandatory to place the target in a magnetic field.
We observe that in the past other experiments,
such as WA17 [41] at CERN, have made use of emulsion stacks
inside an intense magnetic field combined with external tracking.
In this Letter of Intent we will outline the design of such an experiment.