Between St. Andrews Lake and High Falls Lake
Tom Thomson painted both The Waterfall (sketch for the full sized painting Woodland Waterfall) and Autumn, Petawawa along the Barron River (South Branch of the Petawawa River) between St. Andrews Lake and High Falls Lake. This area is two to three hours from the Out-Side-In by canoe. We believe that in Thomson's day, it was also accessible by tote road.
The location of The Waterfall (just upstream of where the river flows into High Falls Lake) is well known. However forest regeneration and the decay of logging works that were active in Thomson's time, preclude getting an exact photographic match to the sketch.
The title of Autumn, Petawawa may be a misnomer because it is possible that it is a spring scene as evidenced by the water flow. The location is the pond a short distance upstream from the waterfall.
We suspect that the sketches Rapids on Muskoka River and Spring Foliage on the Muskoka River may have been painted along this stretch of the Barron River and that the sketch Landscape, Sunset is from nearby. However, the first too are too generic for any definite positive identification and the latter view no longer exists because of obstructing forest growth.
In searching for Tom Thomson sketching locations, we sometimes wondered:
Where are the missing waterfall paintings? Whatever happened to his sketch of High Falls? Of Carcajou Falls? Could Thomson have painted in this vicinity and not have painted these impressive scenes?

(photo by Bob: 2006-05-09 - explore)
High Falls on the Stratton Lake outlet of the Barron River.

(photo by Bob: 2010-07-02 - map - explore)
Carcajou Falls where Carcajou Creek drops into Carcajou Bay on Grand Lake.
The answer might be that he didn't paint those scenes because in 1916 they didn't exist. The Algonquin Park that we see today is not the Algonquin Park that Thomson saw. Logging activity at the time made significant changes to the landscape, both due to removal of trees and to the structures built to make transport of logs feasible. In the 100 years since, the vegetation has grown significantly, and most of the logging structures have decayed away.
Thomson painted relatively few waterfalls because, in his day, undisturbed waterfalls in Algonquin Park were rare. Essentially any watercourse of consequence was used for the transport of logs, and many waterfalls were altered or eliminated by the construction of dams and chutes. There is no sketch of Carcajou Falls because it was most likely hidden behind a logging dam or the view was obstructed by a chute.
There is no sketch of High Falls because it may have been essentially dry. The transport of logs out of Stratton Lake was via St. Andrews Lake and then to High Falls Lake; it was not over High Falls. It seems likely that the High Falls outlet of Stratton Lake was dammed or at least had the flow impeded. The evidence of the river improvements to facilitate this route is still visible today (although we have yet to find definitive evidence of a dam at the outlet of Stratton Lake).

(photo by Bob: 2012-08-08 - map - explore)
The rock chute from Stratton Lake into St. Andrews Lake at low water under conditions of essentially zero flow during the very dry summer of 2012. This stream bed is obviously not natural and has been cleared of obstructions to facilitate the passage of logs.

(photo by Diana: 2007-09-19 - explore)
Remains of a logging chute on the Barron River at the outlet of St. Andrews Lake. Possibly there was a wooden structure supported by these rock walls. Certainly, when the logs were being run in the spring, the water flow would have been much higher.
Autumn, Petawawa


(photo by Bob: 2012-05-07 - map - explore)
The pond on the Barron River between St. Andrews Lake and High Falls Lake upstream of the waterfall. We believe this to be the location of the sketch Autumn, Petawawa. This photo has been stretched vertically by 30%.
The logs would have been collected in the pond before being fed down the narrow log chute around the falls, one log at a time. It is not clear whether the water level in this pond was maintained by a dam at the foot of the pond or not. There is no evidence of a dam remaining today, though there are rock cribs near the downstream end of the pond.
The sketch Autumn, Petawawa depicts a substantial water flow into the pond. This might indicate that this is actually an early season painting. (The strong red hues are indicative but not definitive that it is autumn). In the current era, by midsummer, the flow here is low; the white water is minimal. This would probably be true even with the extra water diverted from the High Falls outlet of Stratton Lake. However, the dam — no longer existing — that controlled the flow out of Grand Lake might have contributed to maintaining the flow later in the season.

(photo by Bob: 2015-05-06 - map - explore)
Remains of a Rock Crib. This is between St. Andrews Lake and High Falls Lake and is upstream of the falls and at the bottom end of the pond. There is another crib on the far shore. It is likely that the logs were boomed here before being fed down the narrow log chute along the far shore and around the falls — one log at a time. There may also have been a dam here, but if so, it was probably only high enough to divert sufficient water into the chute and to keep the pond filled. We know from the sketch, The Waterfall, that most of the flow went over the falls.
The log chute around the falls is gone but some traces remain.
The Waterfall, sketch for Woodland Waterfall

In this case, the waterfall was not drowned by a logging dam, but instead a log chute was constructed on the far side of the river to divert the logs around the waterfall. Thomson does not include the chute in his painting — it is artfully hidden behind the green foliage — but there is visible evidence of it in the painting. It is overflow from the chute that is responsible for the minor flow, visible at the top of the main drop, flowing towards the viewer. This aspect of the flow no longer exists. The main flow of the river is depicted as a thin white line — a few white dots, actually — through the trees in the top right of the sketch.

(photo by Bob: 2015-05-06 - map - explore)
These are the falls depicted in the sketch, The Waterfall. However, the photograph does not correctly capture the angle of view of the sketch. The sketch shows the main flow of the river above the falls (the thin white line through the trees). This indicates that the sketch was done from a position at about the same elevation as the top of the falls (approaching a side view), not from the bottom of the falls where this photograph was taken. However, the modern falls are largely obscured by vegetation at the top and thus the angle of view of the sketch is no longer available.
In the studio painting, (see the Tom Thomson Catalogue Raisonné entry for Woodland Waterfall) the thin white line through the trees does not appear and the overflow from the timber chute has been enhanced to make it appear as the main flow of the river. Thus if you were trying to locate this waterfall based on the studio painting rather than the sketch, you would be looking for a waterfall where the river executed a right angle turn at the brink of the falls and you might reject this location. Bob did for a long time until Diana set him straight.

(photo by Bob: 2015-05-06 - map - explore)
This photo illustrates the difficulty of getting a modern photograph of the exact scene depicted in the sketch. Here Diana is comparing a copy of the sketch with the waterfall from our current best estimate of where Thomson sat to execute his sketch. The minor flow is just visible in the near mid distance with the major flow a little further back. But clearly the modern view is completely obstructed.

(photo by Diana: 2015-05-06 - map - explore)
Diana had to move even to take even this obstructed view of the falls. It has the correct elements although not matching in exact detail. That is, there is a minor flow entering the picture from the right at a seemingly lower elevation while the major flow in the background is visible as a thin white line through the trees.
It seems that the flow was higher when Thomson painted this scene in 1916 than it was in early May 2015.
Spring Foliage on the Muskoka River

With this sketch, there is not much to go on. But the layering of the far shore rock face and overhanging cedar branches matches the shoreline just downstream of the falls. For many years, the sketch was titled Rushing Stream

(photo by Bob: 2015-05-06 - map - explore)
Here is a photo representative of that area but was not taken with the intent of matching the sketch. However, the cedar branches and the dip of the bedrock match the sketch.
It is interesting to compare Thomson's painting of the rock face on the far shore in this sketch, Spring Foliage on the Muskoka River, with that in The Waterfall and with the finished painting Woodland Waterfall. We suggest that this similarity supports our hypothesis that Spring Foliage on the Muskoka River was painted just downstream of The Waterfall.

A composite of Spring Foliage on the Muskoka River, The Waterfall, Woodland Waterfall, and a photo of the rock face below the waterfall.
Rapids on the Muskoka River

This is a pretty generic small-river, white-water scene. It may well be the Muskoka River. But visually, it is also consistent with the Barron River between St. Andrews Lake and the pond.

(photo by Diana: 2015-05-06 - map - explore)
The Barron River just upstream of where it drops into the pond. Suggestive but not conclusive. Another generic northern river in the spring. (There may have been a log chute here in Thomson's day.)
Landscape, Sunset

What we think this sketch depicts is the rocky bowl surrounding High Falls Pond, with the entire area largely denuded of trees. The sketching location is between the Barron River and High Falls. High Falls is falling into High Falls Pond in the center of the sketch. But it doesn't really show up because either there was minimal flow due to that outlet of Stratton Lake being dammed, or because it is in shadow, or both. The location of the "water slide" might be just visible (or just hidden) by the middle-distance rising ground on the left. The far blue hills are those on the south shore of Stratton Lake.

(photo by Diana: 2015-05-06 - map - explore)
We have searched this area for a matching view, but it is a hopeless task given the current tree cover. Here's an attempt at photographing the described scene in the modern era.
Notes
Officially, Carcajou Falls at the top of Carcajou Bay on Grand Lake are named High Falls. But there are too many "High Falls" in Algonquin Park — this one on Carcajou Creek, one on the Barron River, one on the Nipissing River, one on the Bonnechere River, one on the York River, and perhaps more. We prefer Carcajou Falls as a better name for these particular falls.
The assertion that there was a dam at the foot of Stratton Lake is speculative, though it would help explain the rate of flow of the river in the sketch Autumn, Petawawa, especially if indeed, it is an autumn picture. However, none of the sketch locations identified on this page depend upon this assertion.
While we have not (yet) found evidence of a dam at the High Falls outlet of Stratton Lake, in the spring of 2015 we perhaps found evidence of a crossing at this point and an old road. We need to revisit this location for further investigation. Also, there was an old lumber camp located in the near vicinity (at the group of four backpacking campsites).