Against "Philosophy" the OED quotes "seeking after wisdom or knowledge2. Well I've certainly been doing that since I was unable to accept the J10 rear splashers looked like those in pictures of a J10.

Because of my modelling philosophy which has developed into "copy everything exactly to a ratio of 76.2:1 if physics and engineering limitations will allow", I've run into problems. Any genuine finescale modeller worth his salt knows that physics in general does not allow total accuracy in such small models.

The problem from the beginning has been the lack of a reliable drawing of a J10 as of around 1946-8.

Way back when I made my very first scratch-built model, an LNER C13 with a plastikard body and a modified Triang chassis, now pushing toward 50 years ago, I had already made the decision that I was a copier and that my skills would meet a target of +/_ 0.5mm accurate reproduction, as long as I had a decent drawing of what I was making.

I was naive enough to consider that any drawing published in one of our magazines would be accurate. Absolute nonsense of course, and much the same applies to purchased MR drawings. Indeed I think that I can claim fairly certainly that on occasions where I have checked drawings against some known dimensions or, years ago against a prototype, I have found errors. The exception has been 'Isinglass' drawings but John had not done the J10 when he died and I don't think the present owners intend to expand the drawing list. Just once I thought John had erred but in fact it proved to be every other copy of an LNER O4 drawing that had the error, the Isinglass drawing was quite correct.

So all my problems with the J10 splashers can be blamed on 'Isinglass', - they haven't drawn the J10! ~(Just in case the reader is too serious minded, my tongue is very firmly in my cheek).

I have drawn it myself and proved how very easy it is to make slight mistakes; because full information is not available, it is necessary to make measurements and assumptions that any engineering training warns you against. 'Scaling' or taking measurements from a drawing is a quite definite no-no in engineering, particularly precision engineering. Why?

Because drawn lines have thickness; - do you measure to one of the edges of a line or its centre?

Because the original draughtsman may not have been so skilled as one would wish. (Computers are more accurate but the LNER did not have one let alone the MS&LR!)

Because paper is not stable, so the print may end up a different size to the original drawing. In our case we often use a photocopy of a photocopy of a ....... etc, and you can almost guarantee that your print is quite a different scale than the original.

My J10 drawing was made around 2000 and I used a print from a Bayer-Peacock drawing dated 1898. Both Pollitt and Robinson made mods. to these locos and Gresley came along after those two, so my requirement for a 1946-8 version would almost certainly have some changes. I worked out a multiplying factor for my print using the known diameter of the driving wheels, but I did my very best only to use dimensions which were actually indicated by the original draughtsman. None-the-less some 'scaling' had to be resorted to. I am experienced enough in this area and I was well aware that my drawing was of doubtful accuracy but it was based on the best information available. I have found no dimensional information about the modifications made by the later designers but I was aware that the width (length?) of the cab side sheets was increased and I made some estimations here based on the J11 & O4 cab side sheets.

Since my last posting here, I spent all of Saturday morning,- about four hours, - re-checking my drawing against the B-P print. Then I scoured the relevant Yeadon and RCTS 'green bible', looking for full-side-on pictures, and settled on two in Yeadon (p31 & p43). A scale was calculated for each of these and various dimensions measured on the drawing and on each picture, resulting in three measurements for any given thing or position. As an example the width of the side sheet came out as: B-P 22.7mm; p31 pic. 23.2mm; p43 pic. 23mm. The mean of those three comes out at 22.9666, so my model side sheet at 23mm is acceptable.

My thinking now went along the lines of: - "the front edge of the side sheet must be too far back from the rear wheel centre line, hence the need for an enlarged splasher". The mean of three measurements method was again applied & resulted in 7mm. The model's equivalent was 6mm. Was the chassis wrongly fixed to the body? Check the dimension from buffer beam to centre-line of front axle; - mean value 22.5, model currently at 21.5mm. Simple error; the chassis is 1mm too far towards the front, move it back 1mm. Good, because the chassis fixing could easily be modified to do that.

Sunday morning found me down at the 'Tank' with the simple task of modifying the chassis fixing hole. Well the task is easy but the end result puts the splashers 1mm back of their wheel centre line and the whole assembly looks idiotic!

Option one; - move the splashers. Not possible because the slots in the running plate below the splashers are also 1mm off centre line!

There is only one sensible answer. I have to remove the splashers, valances, buffer beam, drag plate and the cab from this running plate and make a new one. I'm back where I was about three months ago and not very happy. However I know that I will be happy when the J10 is running on the layout. Indeed I'm fairly happy now because I've found out what was wrong.

My year 2000 drawing needs some further correction!

Funny hobby railway modelling

Dave