Saturday 9 May 2015

Guest Contributor – Robin N Campbell


Introduction

I have approached a number of people to write articles, but if readers would like to contribute an article please contact me. The only two stipulations I make are that the article has to be hill related and that I don't end up in court through its publication! Otherwise the choice of subject matter is down to the Guest Contributor.

Guest Contributor – Robin N Campbell 


Robin Campbell, a few years ago, with two good friends

Originally published in The Munro Society Journal No 2 2010 (Photos: copyright SMC image collection)

Corbett’s Criterion
Robin N Campbell

John Rooke Corbett died in 1949 after years of illness dating back to 1943, when he had suffered a heart attack. His sister Catherine subsequently passed over a list of Scottish hills 2,500 ft. and under 3,000 ft. in height to the Scottish Mountaineering Club, and of course it is these hills that have come to be known as the Corbetts. The list contained no information about the criteria for inclusion, beyond the obvious one specified by the title. Indeed John Dow, in his introduction of the first publication of Corbett’s list, seemed somewhat baffled. He described the list as ‘incomplete’ and commented that ‘numerous heights of equal merit to those listed have not been shown’. He argued, however, for publication in this incomplete form ‘as a memorial to a personality genuine and likeable, rather than as an authoritative document’. ‘Perhaps,’ he added, ‘at some later date, we may see a revised Edition, comprehensive of every 2,500-foot summit.’ So Dow was unable to understand at that time why some 2,500 hills had been included while others were left out.

Corbett, on right, on the summit of Cona Mheall, 1938
However, when Corbett’s list was published as part of Munro’s Tables, Dow wrote, ‘There was no indication in Corbett’s papers as to the criterion he adopted in listing the heights included, but it seems clear that his only test was a re-ascent of 500 feet on all sides to every point admitted, no account being taken of distance or difficulty. No detailed check has been made, but the 500 feet qualification has obviously been exhaustively applied and rigidly adhered to’. Dow’s inferred criterion ‘with a re-ascent of 500 feet on all sides’ was then, and henceforth, added to the list’s title.

‘Seems clear’, and ‘no detailed check was made’ are phrases that are apt to excite the scholar, so there is some room here to wonder whether Dow’s inference was the correct one. Certainly the current list of Corbetts results from the exhaustive and rigid application of the 500-foot rule, but there have been many changes in the list of Corbetts since 1952. Had the SMC perhaps changed Corbetts’s criterion as well as his List? I believe that it did.

Short of a detailed examination of every 2,500 hill on and off Corbett’s list, there are quicker methods of putting the 500-foot rule to the test. Six hills were removed from the list in 1981 and two more in 1984 on the grounds of inadequate drop. These are shown in Table 1 below.

Name
Location
GridRef
D&H
Year
Map
Cont.
Ht.
Drop
Beinn a’ Chumhain
S. of Ben Alder
NN462710
4B
1981
48
10
2958
482
Sgor Mor
NE. of An Socach, Glen Clunie
NO115825
6B
1981
49
11
2908

Meall Uaine
S. of Spital  of Glenshee
6B
1981
49
10
2600
476
Carn Easgann Bana
E. of Fort Augustus
NH485063
9B
1981
42
10
2554
463
Druim Garbh
W. of Sgurr Domhnuill
NM881683
18B
1981

47
10
2637
466
Beinn an Tuim
N. of Glenfinnan
NM929835
10D
1981
47
10
2656
472
Kirriereoch Hill
N. of Merrick
NX420870
27B
1984
83, 87
10
2565

Cramallt Craig
S. of Broad Law
NT168247
28B
1984
80
11
2723


















Table 1. Corbetts removed since 1952 because  of drop < 500 ft. Final column shows drop [= re-ascent] computed from Six-inch spot heights where these are available. ‘D&H’ gives hill sections in Corbett Tops and Corbetteers. Alan Dawson & Dave Hewitt. TACit Press, 1999. ‘Year’ refers to the edition of Munro’s Tables in which the hill was removed. ‘Map’ is the Popular One-Inch sheet. ‘Cont.’ is the number of contour lines crossed from defining col to summit on the identified map, and ‘Ht.’ is the height given on that map.


It is obviously important, here and elsewhere in this article, to consider only map data which was available to Corbett. I have used the ‘zoomable’ digital map of Scotland accessible through the National Library website. Very conveniently, the zoom passes through the ‘Popular’ One-Inch Survey to the Six-Inch, and these are the same maps available to Corbett. I presume that I have to thank Ordnance Survey copyright restrictions for this splendid historian’s tool, as well as the National Library.

What is immediately noticeable is that for six of these hills only ten contour lines intervene between col and summit. Ten contour lines enclose only 450 feet. If Corbett intended to include only hills with a 500-foot drop, then further data was needed for such hills, especially since five of them had ‘surpluses’ over the top contour of only a few feet. In fact, such data was available to him from the Six-Inch maps in five cases, and this data shows (final column of Table 1) that each of the five had drops comfortably below 500 feet. Even without the Six-Inch data, contour-interpolated col estimates may be taken from the One-Inch map. Since this adds only 25 feet to the summit surplus, we can see that we still have five hills with estimated drops comfortably below 500 feet. So these hills were removed by SMC not because of revised O.S. measurements, but because they didn’t meet the criterion of 500 feet of drop, which Dow and perhaps Jim Donaldson had ‘deduced’ on Corbett’s behalf.

If Corbett consulted the Six-Inch maps, or made use of col estimates from One-Inch contours, then we may safely conclude that the SMC wise men were in error, and that Corbett intended as criterion ‘drop > 450 feet’. If Corbett didn’t consult the Six-Inch maps or use estimated col heights, then he was obviously content to include any hill with 10 intervening contour lines. In either case, regardless of what Corbett may or may not have had in mind, the criterion that defined his list of hills in its original form was ‘drop > 450 feet’.

SMC New Year meet 1932, Newtonmore.  Corbett 2nd from right.  Front row from the left:  Archie Robertson, James Gall Inglis, John Dow
Cook’s Cairn (Glenfiddich) was removed in 1981 on the grounds of inadequate height. Corbett’s One-Inch map indicated a summit at 2541ft. However, this was based on a mistranscription of a Six-Inch spot height of 2341ft, situated 200m. SSW of the actual summit of 2478ft. So, we have some evidence that Corbett didn’t look systematically at Six-Inch data. However, before considering the question of whether Corbett consulted any data other than the count of contours, it is essential to settle a more basic question. We need to know (a) whether he included all hills with 10 or more intervening contours, and (b) whether he left out all hills with less than 10 intervening contours. As to (a), if that was his criterion, then clearly they should all be included. As to (b) if he included any hills with just 9, then ‘10 intervening contours’ was obviously not his criterion.

To see whether he included all hills with 10 intervening contours it is necessary to examine those hills with substantial drops which were not included either by Corbett or by succeeding editors of his List. Dawson & Hewitt’s excellent booklet (page 27) provides a list of suitable candidates in descending order of drop. I have examined the 27 hills in their list with drops in excess of 134m. (440ft.). None have drops with more than 9 intervening contours.

The second of these 27 hills, Marg na Craige (Laggan), deserves comment. The Marg is of course interesting in every way: any sensible classification system would grant it independent status, since it is the summit (2731ft.) of an extensive range. But there's a puzzle about the Popular map markings. There's a spot height of 2255ft. at the col, giving a drop of 476ft. But the contour lines imply that the col is between 2300 and 2350 feet, hence there are only 8 intervening contours. The Marg's surplus over the last contour is 31, leaving a drop of 445ft. to find to the spot height, which is just short of 10 contours. So even if we take the 2255ft. spot height as definitive data, it's still a 9 contour hill.

To see whether Corbett excluded all hills with less than 10 intervening contours, it is first of all necessary to examine those hills in his List with low drops. Leaving aside the hills of Table 1, which are already established as having 10 or more contours of drop, it is not too difficult to examine relevant others, since surprisingly few Corbetts have drops in the vicinity of 500 feet. I looked at Cam Chreag (Lyon), Meallan nan Uan (Strathconon), Geal Charn (Lochy), Carn Ealasaid (Cock Road) and Morrone (Braemar): all have at least 10 intervening contours. Secondly, there are a few hills added since 1952 for various reasons. Table 2 below shows those hills added to Corbett’s List by SMC since 1952, on account of having a drop > 500 feet. Since Corbett didn’t include these hills, none should have 10 or more contours of drop.


Name
Location
GridRef
D&H
Year
Map
Cont.
Ht.
Drop
Beinn Each
S. of Stuc a’ Chroin
NN602158
1A
1984
62
9
2660

Sron a' Choire Chnapanich
W. of Stuchd an Lochain
NN456453
2A
1984
55
4
2250c.

Meall na Meoig (Beinn Pharlagain)
S. of Sgor Gaibhre
NN448642
4B
1981
48
9
2836

Cam Chreag
W. of Creag Mhor, Glen Lochay
NN375346
2B
1981
55
9
2887
485
The Sow of Atholl
W. of Drumochter
NN624741
5A
1981
48
7
2500c.

Sgurr Cos na Breachd-Laoigh
N. of Upper Glendessary
NM948946
10B
1981
41
8
2600c.

Sgurr Gaorsaic
W. of Sgurr nan Ceathramhnan
NH036219
11A
1981
36
8
2600c.

Sail Ghorm
W. peak of Quinag
NC198304
16E
1981
15
6
2551


Table 2. Corbetts added since 1952 because drop > 500 feet. For column headings, see Table 1

Clearly no hill in Table 2 had more than 9 intervening contours in 1952, so if Corbett’s rule was ’10 intervening contours’, he should have left these hills out, and he did. On the other hand if his rule was ‘drop > 450’ and he had consulted other data, then he would have included Meall na Meoig (contour-interpolated col estimate of 2375ft.) and Cam Chreag (Six-Inch col spot of 2402ft.; or contour-interpolated col estimate of 2425ft.), and he didn’t. Sron a’ Choire Chnapanich deserves a mention as being possibly the worst-mapped Scottish mountain. Its summit is almost 500 feet higher than the 1952 height.

I believe that all relevant hills have now been considered and it can be stated with confidence that Corbett’s criterion for inclusion in his List was, besides the height criterion, a drop criterion of at least 10 contour lines intervening between col and summit. That is, his List was a list of Scottish 2,500 foot hills with a re-ascent of 450 feet on all sides. There is no evidence that he consulted other available map data, and considerable indirect evidence that he didn’t look at Six-Inch maps or use contour-interpreted col heights. Had he done so, his List would have been a different one, regardless of whether he had a criterion of 450 or 500 feet of drop in mind, since either there would have been some 9-contour hills that met the 450 foot drop criterion, and were therefore included, or there would have been some10-contour hills that failed to meet the 500 foot drop criterion, and were therefore left out. It is surprising that he ignored such data: he was a mathematician and land tax assessor and thoroughly familiar with the use of such data by Munro and others. Perhaps he had intended to, but failing health prevented it.

As a consequence of the observations in this article, and if it is desired to keep faith with Corbett’s intentions, the List of Corbetts should be revised to include the 21 new and reinstated hills with current drops between 500 feet (152.4m.) and 450 feet (137.2m.). These are the first 21 hills in the list on page 27 of Dawson & Hewitt’s booklet.

Finally, although I will not dwell on it here, I should state that it is my belief that the identification of hills as outstanding in some way should always be made by considering some function of distance and drop, as was done by Hugh Munro, Percy Donald, William Docharty and Fiona Graham, and not by drop alone – as has been done by Corbett and Alan Dawson, and by various tabulators of English and Welsh hills. The use of drop alone has led to absurdities such as the vast range south of Glen Lochsie and west of Glenshee – outstanding in every sense – being treated as a mere appendage of Glas Tulaichean, and to the fine outstanding mountains of Quinag and Sgurr Dhomhnuill being split into numerous hills, each with tiny ‘footprints’. The method of Munro is required, and should be insisted upon by all students of mountain form.





Originally published in the SMC Journal (Photos: copyright SMC image collection)

The following article first appeared in the Scottish Mountaineering Club 
Journal (SMCJ, 202/41, 341–2) Copyright © 2011, and is reproduced here 
with permission.



Corbett’s Drop Criterion[1]

Robin N Campbell

Since its appearance in Munro’s Tables in 1953, Corbett’s ‘List of Scottish Hills 2,500 and under 3,000 feet in Height’ has been announced as selected by the criterion that each listed hill has a drop of 500 feet on all sides. However, this criterion was never stated by Corbett[2]. Instead, the editors John Dow and Jim Donaldson[3] deduced that this was Corbett’s criterion, which ‘has obviously been exhaustively applied and rigidly adhered to’, said Dow. I believe that this deduction was incorrect.

The ‘zoomable’ digital map of Scotland accessible through the National Library website provides a convenient means of examining the hills included by Corbett, and those left out, in relation to the map data available at the time. The zoom passes through the ‘Popular’ One-Inch Survey to the Six-Inch Survey, and these are the same maps that were available to Corbett. I presume that I have to thank Ordnance Survey copyright restrictions, as well as the National Library, for access to this useful historical tool.

An examination of the hills included in the 1952 List in relation to the Popular Map shows that all of them had at least 10 contour lines intervening between the drop-determining col and the summit. An examination of those hills excluded from the 1952 List shows that all of these had 9 or fewer contour lines between col and summit. Corbett’s drop criterion was therefore ‘at least 10 contour lines intervening between drop-determining col and summit’, and it is this criterion which was ‘exhaustively applied and rigidly adhered to’. The Popular Map plotted contour lines at 50-foot intervals, and so 10 contour lines enclose 450 feet. Effectively, then, Corbett’s drop criterion was ‘450 feet or more’.

SMC Easter Meet 1931, Tomdoun.  Corbett centre in big hat
Naturally, the application of this criterion will result in the inclusion of some hills with between 450ft. and 500ft. of drop. Five 10-contour hills were removed from the List in 1981 and a sixth in 1984, because they failed to meet the drop criterion of 500 feet. But these deficiencies of drop did not result from any relevant change in the mapping of these hills. Data was available in 1952 (though it is unlikely that Corbett consulted it) to associate these 6 hills with drops between 450ft. and 500ft. Dow and Donaldson, had they noticed this in 1952, would have concluded that Corbett’s drop criterion was 450 feet, rather than 500 feet. If Corbett had consulted Six-Inch col heights, and/or contour-interpolated One-Inch col estimates, his list would have been a different one, whatever height-specific drop criterion he had in mind, if any. In fact, there is no evidence that he looked at any data except the summit heights and the contour markings of the Popular Map.

If Corbett had consulted Six-Inch col heights, and/or contour-interpolated One-Inch col estimates, his list would have been a different one, whatever height-specific drop criterion he had in mind, if any. In fact, there is no evidence that he looked at any data except the summit heights and the contour markings of the Popular Map.

As a consequence of these observations, and if it is desired to keep faith with Corbett’s intentions, the List of Corbetts should be revised to include 21 new or reinstated hills with current drops between 500 feet (152.4m.) and 450 feet (137.2m.). These are the first 21 hills in the list on page 27 of Alan Dawson & Dave Hewitt’s booklet Corbett Tops and Corbetteers (TACit Press, 1999). A slightly different possibility, less precise but perhaps more in keeping with Corbett’s actual criterion, would be to add those Corbett Tops which exhibit at least 15 contour lines intervening between drop-defining col and summit on current maps. Since 15 ten-metre lines enclose 459ft., this is effectively a slightly more severe criterion than Corbett’s. Its application would add only 19 new or reinstated Corbetts.


1  A longer version of this note, with more detailed argument and discussion, appeared as ‘Corbett’s Criterion’ in The Munro Society Journal, No. 2, 2010.
2  Dow’s introduction to the first published version in 1952 makes this clear (J, 25, 45-52).
3 Jim Barton (J, 41, 36) showed that it was E.W. Hodge who originally made this deduction.





 
Page 27 of Alan Dawson & Dave Hewitt’s booklet Corbett Tops and Corbetteers (TACit Press, 1999) as mentioned in above paragraph, appears below.




1 comment:

Henry M said...

First question is, did Corbett regard the number of contour lines as particularly important, or whether they were a convenient way of estimating re-ascent? Spot heights are definite (if necessary by special survey). Contour lines are arbitrary, change if you change the contour interval.
Second, did Corbett want to be sure he didn't miss a potential candidate - it would be like including the sub-Marilyns and Marilyns in a single list.