 So the root time method involves taking the data that shows the change in sample thickness and plotting that against the root of time. So if you have sample thickness on your y-axis, we call that h, sample thickness, and we plot that against root time on your x-axis, you would usually get a line that looks something like this. So to get your CV value from that, you need to do a number of different steps. So the first step you need to do is take the points at which this line is linear. So it has an initial rapid compression as the load increment starts, but eventually it reaches a linear proportion or relatively linear proportion of the curve before reaching some sort of equilibrium. So we take the linear proportion of that curve and if you then backcast that linear proportion to the y-axis, so if I find the gradient of that in a draw line, that crosses the y-axis and that crosses at point A, as an intercepted point A. If I take the gradient of that line and if I then divide, I'll write it on the other side. If I take the gradient of that line and I divide it by 1.15, I'll get a new gradient. So let's call that, let's call this M1 and M2. So my new gradient M2 will be 1.15 times shallower than the gradient of this line. Okay, so if I draw a new line with the equation where M2 is now my gradient, but that has A as the intercept, so the equation of my line will be y equals M2x plus A. So if I take, that is my new line and I draw that onto my diagram, a line that looks something like this with a shallower gradient. So the point at which this new line crosses my compression curve, so the original data, the point at which it crosses the curve, which is here, let's call that point B. That's equivalent to my T90 value. So the time for 90% of consolidation, so well that'll be, I mean it will still be in root 90, so we need to square it to get it into just T90. So square it, that'll be T90. And I know that if T90, my TV value will be equal to 0.848. And I know that from either an equation or a table. So it's T90, I know my TV value is at 0.848. Now I can then go back to the equation that relates TV to CVT over D squared. Now in an odometer test, I know what the sample thickness is. I know the sample thickness and I know that it drains from both sides. So I can take the sample thickness and divide it by 2 and I can get my drainage pathway. So I know that from the odometer test. So I take half my initial sample thickness. The T value is this value here, T90. So I can use my T90 and I can use my TV value here. So this is, would be equal to 0.848. CVT90 over D squared. So I know the T90, I've just worked that out from my graph here. I know the D value from the odometer test and I know the TV value. So the only thing missing is my CV value. So if I rearrange this equation to make it the subject of, make CV the subject, I should be able to work out what my CV value is from this information. So in an odometer test, we do a number of load increments. So we start with an initial sample thickness, which is what the y-axis is, and time here. We put a load increment on and we wait until it reaches an equilibrium. Usually it's 24 hours between loading increments. So we wait for it to reach equilibrium and then we add another load and then another load. And you can see that actually you can do a root time method on all of these steps. And that will give you a different CV value for each increment. So we know that CV is proportional to the change in MV. And we know that MV changes when we're increasing our load increments. So if that's the case, then we would expect to get a change in CV value for each of these loading increments. Another interesting thing to point out from the new odometer test is if we use the root time method to get a CV value and we can also derive an MV value from this test. If we go back to the equation that relates CV to MV, if we can generate CV and MV from this test, the only unknown within this equation is permeability. So another interesting aspect is that we could use the odometer data to derive a term for permeability. And that adds just an extra bit of value to this test. And it's something that you should work out if you're doing it.