Notes

Small nu writeup: keep all the plots but move them to a folder where you keep everything - i.e. away from actual writeups. Decide what you want to convey in the section. You want an example of the shape and meniscus (asymptotics vs numerics), loglog scale plot of |t_end| against |nu| for different values of w and x_0, log scale plot of error delta t (but multiplied thru by nu^2 so it's the relative error rather than absolute) for various different values of w - should see straight lines w slope 1. Repeat for delta, with inset showing the errors (just mention in passing) - depends only on w (point this out - expected?). Point out the CURIOSITY that at small w, then wt_{end} looks to be independent of w, but don't persue it (small w may be contradictory to earlier asymptotic expansions).
Small w: proceed with what you have been doing, write up neatly with rescaled time and X. Matching comes from writing the lower shapes using x = x_l + w*X, expand and compare. Ian suspects that H = const(t) at leading order - i.e. there's an initial squeezing and then some motion around this, expand H = H0(t) + w^a H1(X,T) etc. Make plot of when it initially moves downwards for the various w data sets you already have. Look at a few numerics examples to see if you can tell what's happening as w shrinks - timescale w^-5?
Transfer: Get a plan in place. Lit Review (what do you want to cover), intro (set the scene + motivation (expts!)) -> model -> non dim -> asymptotics -> 'extensions'. Extensions: where are you going to go with it? Future work ideas- Hele Shaw cell, hysteresis (say what can happen and prove there is actually something to do!), when do equilibria exist (and can I find these with numerics)
Experiments - different viscosity silicon oils. Check timescales vary with viscosity - use in intro.