Is there a better way to do this?

[throbscottle]

Here is a two transistor SCR version that is a little bit easier to turn OFF. Q2 would be your drive transistor and should be rated for whatever load you are using. Q1 is low current and can be just about any PNP transistor. D1 and R5 are essentially your load.
**broken link removed**

Nice idea but I need SW2 to turn it off even when SW1 is still closed.
Been ill. Not had chance to try anything!

 

[throbscottle]

Dammit. Whilst testing the 2 transistor SCR (which works well on it's own btw) I remembered that part of the original plan had been the circuit needs to cancel it's output when the input condition ends, not just when the reset button is pressed! This works in the original ugly design but when the input is turned into a momentary event the function is lost.
I think the dual D might still do it with an extra cap between clock and reset.

 

[eTech]

Hello,

Here's a one shot version of the circuit I posted earlier.
The timing can be canceled by the clear button.

 

[throbscottle]

Hah! Done it.
I tried a 4013 and found I can't get sufficient logic levels with a LDR as the input, so played with the original design, discovered some things I didn't know about 555 timers, took a lot of stuff out.

So now U1 works the same as before, but U2 instead of the toggle function acts as an inverse output set/reset latch with Q1 providing the negative going trigger (to reset it high) when the input event finishes. Pressing the cancel switch sets the output low and resets U1. C2 stops a blip happening when the input event finishes, if the cancel switch has been pressed.

(edit) So I think I'm going to use this one, with a 556. Thanks very much for all the suggestions, everybody!