41. On the forest stations must be built as required-Overseer's headquarters and offices, nurseries and experiment areas, camps, paddocks and pounds, and look-outs, &c. Standard specifications should be drawn up by the Branch of Forest Engineering.

42. Logging railroads and chutes should be laid out into bodies of otherwise inaccessible timber.

43. Telephone lines may have to be built in some instances, to connect forest stations to settlements.

All this work is within the especial province of the proposed branch of Forest Engineering.

Its execution should be under the direction of the forest supervisor and the foremanship of the forest overseer. Under this organisation, day labour should be most appropriate and effective in forest work.

BUILDING A SYSTEM.

The goal of a Forest Service in every aspect of its work should be "100 per cent. Efficiency."

That goal can never be attained by the adoption of a "muddling through" policy. There must be a persistent scientific attack upon the problems to be solved. The importance of detail must be recognised. The principles of efficiency engineering should be incorporated into departmental routine.

There should be an early stocktaking of existing methods-a setting down in writing of the accumulated "head" knowledge and experience of all officers, an inventory of laws, regulations, precedents and procedure. There must be a sorting out and a standardisation. A provisional system must be adopted as a basis for progressive revision.

I suggest

44. The preparation of a State Forest Manual parallel to that of the United State of America Forest Service; to be bound on the loose leaf principle so as to allow of revision by the substitution of new pages for old. 45. The delegation of a special officer to undertake its compilation and progressive revision.

In the meantime, the following suggestions are offered by the United States of America experience with respect to the development of a system for each branch of forestry work.

(1.) Forest Classification.

Timber is an elementary necessity of State. Forest demarcation is a vital issue.

There must be an early and complete stocktaking of the timber resources of each State, with a view to determining what areas should be reserved permanently, what temporarily, and what made available for settlement. Australia is a comparatively timberless country, and it limited forest areas are particularly precious. But the forests generally are situated in the regions of greatest rainfall and accessibility, where settlement congregates. Their possession consequently has been contested fiercely by the opposing interests.

The theory of setting apart the poorer lands for forestry holds to an infinitely less degree here than even in Europe or America, since the climate, configuration, and location are such as to make in Australia even comparatively poor lands desirable in the eyes of the settler and the land speculator.

46. There must be a sufficient forest reservation to supply the material needs of prospective populations and to afford safeguards against possible climatic and waterflow changes.

47. A per capita basis cannot be adopted because the Australian community is a mere nucleus of what it is destined to become.

48. Since most forest lands are more or less suitable for some form of settlement, and the theory of relegation to poor soils will not hold in Australia, an arbitrary forest ratio must be established for each State in framing the forest policy.

49. As far as is at all practicable the Australian ratio should be fixed at not less than 5 per cent. of the total State area. In Europe the proportion of 20 per cent. to 30 per cent. still holds, in despite of teeming populations and intense demand for land, and forest redemption is still taking place. Mr. D. E. Hutchins in his "Discussion of Australian Forestry" proposes 15 per cent.

50. Preliminary forest exploration work should be pushed as far and as quickly as practicable. A cursory examination should be enough to secure a temporary reservation of likely lands pending final demarcation. The Departments of Lands should co-operate in the project.

51. Final demarcation up to the limits of the prescription for the State should be the exclusive function of the Forest Services, and should be undertaken in the course of forest survey and assessment.

52. State Forests should be located as closely as possible to present and prospective markets, because of the heaviness and bulk of the material to be transported therefrom.

53. They should be of sufficient extent to make possible and profitable their management as separate areas. Forest lands are most valuable when handled in large unbroken units.

54. Primarily, they should furnish a soil and climate adapted to the production of marketable timber. If they carry seed trees of the right species, so much to the good; if a mature stand exists, still better; if an abundant young growth, best of all. If the cover furnishes watershed protection, reservation may be justified even in default of other factors. Ruggedness is an argument in favour of retention. Preference should be extended to mountain forests.

55. Areas necessary for administrative purposes should be retained.

56. Administrative purposes include forest stations, plantations and nurseries, mill and camp sites, log depôts, logging roads and chutes, and means of access, look-out points, telephone lines, firebreaks, and every other area needed in the work of the forester, and such areas as are requisite for public camping grounds, health or pleasure resorts, or other purposes for the use of the public.

57. Since Australian forest exploitation depends primarily upon bullock teams, the provision of water supply and pasture, paddocks, pounds, stock driveways and holding grounds, is of especial importance, apart from the question of general forest grazing management.

58. Patches of agricultural or other non-forest land which might otherwise be eliminated, but are so intimately an integral part of the forest that they cannot be segregated without destroying a forest unit, should be retained. A State forest should not be shredded by alienation to a point making protection, general administration, or application of forest management unpractical.

59. Agricultural land which might otherwise be cut out, but carries a heavy stand of timber of such value that the timber speculator in place of the bona fide selector would seek to acquire title, should not be made available until the timber has been removed completely.

60. When a considerable area of forest is clearly of great economic value to the State and an attempt to transform it into an agricultural country might result in irreparable disaster by disturbing a well-established climatic or economic equilibrium, it should be retained. Such an area is the Pilliga Forest of New South Wales.

(2.) Forest Survey and Assessment.

Forest assessment is an operation which must be repeated periodically. Forest survey, however, must be undertaken once and for all. The two projects are inseparable at this stage, and, combined, they represent an undertaking of paramount importance, since it furnishes a foundation upon which must be erected the whole edifice of forest management. A comprehensive scheme of forest survey and assessment should be conceived at the start. It will give the greatest return in the long run.

61. There should be a broad, but elastic, division of the two component parts of (1) survey: and (2) assessment; the first to be directed by the proposed branch of forest engineering, the second by the proposed working plans branch in collaboration.

62. Forest survey should cover primary and secondary control, the objects of which are to provide an accurate surround, and a framework of more or less parallel lines and profiles not over two miles apart.

63. Primary and secondary control may be deemed to be the especial province of the proposed forest surveyors.

64. Tertiary control and assessment would cover the method of strip survey, with the aim to clothe the secondary control framework with topographic detail and to gather working plan data.

65. Asesssment work, including the collection of working plan data, is a matter for the exclusive attention of experts acting under the direction of the Working Plans Branch.

66. Small separate camps are uneconomical, and considerable advantages flow from a combined organisation such as that consisting of a forest surveyor and two assessment parties (a forest assessor and his assistant) working together. Camp transport work is reduced, and one cook suffices for all; camp life is made more congenial and the work is expedited generally.

67. Secondary control lines should be blazed, and marks should be set up as the starting and closing points for horizontal and vertical strip surveys. 68. A camp map should be prepared, on which should be plotted the surround and framework, and all existing surveys.

69. A preliminary subdivision of the area should be worked out on the camp map for purposes of convenient estimating and compiling. The United States of America Forest Service uses the 40-acre lot as a basis.

70. The projected strips should be laid out in pencil on the camp map. The distance between strips should be ten or twenty chains. Circumstances, however, may alter cases.

71. A scale of 20 chains per inch and contours of 25 feet interval should be adopted as standards for the detailed base maps resulting from strip surveys. 72. I recommend the use of the new Bonner Reflecting Abney level and trailer-tape system of tertiary control. The United States of America Forest Service sketching-sheet and transparent duplicate with separate tally sheet. may also be employed with advantage. The tally man should book tallies instead of calling out to the assessor. Perfect synchronisation is requisite if speed is to be attained.

73. The survey season should be fixed to suit local climatic conditions as far as possible; ordinarily, it should be in the autumn to spring months. The severe summer months might be devoted to office work or to details of study and investigation at the Forest Institute and Experiment Stations, or holidays.

74. For the proper assembling of data, four standard maps should be adopted, viz., (1) The base map, showing topography and "culture"; (2) the silvical map, showing forest types and age classes; (3) the timber map, showing timber stands and assessment estimates; and (4) the forest organisation map, showing present and projected improvements and subdivisions.

75. A State forest atlas of a standard size (say, 18 inches by 21 inches) should be established at head offices, with district forest atlases at each district office. In these atlases should be assembled systematically the prepared maps and collected figures resulting from strip survey. An index map on a scale of 80 chains to the inch should be included.

For map-filing purposes, the adoption of the Beck vertical map file and envelopes is recommended.

76. The American alpha-numeric symbolisation scheme to denote mixed stands should be adapted for Australian use.

(3.) Silviculture and Forest Management.

Silviculture is the handling of woodlands in accordance with the various silvicultural characteristics and requirements of the contained species; in other words, the incorporation into forest treatment of the results of the forest research to be carried out at the forest. institute, experiment stations, and demonstration forests, and by forest officers generally. Its objective is the attainment of the normal or perfect forest.

Necessarily, it cannot come in its fulness forthwith; the unfolding will be gradual. But its application begins with exploitation, which, silviculturally speaking, is only a means to an end, that of normal regeneration.

77. The welfare of the stand must be the chief consideration in permitting cutting operations, which should be conducted so as to afford the maximum opportunity for natural regeneration.

78. There must be an insistence upon conservative methods of timbergetting (low stumps, high tops, sawing instead of axeing, &c.), not only upon grounds of economy, but in order to reduce the accumulation of debris and improve the chances for reproduction; where necessary, slash must be stacked and burnt, or scattered; dead, defective, and damaged trees must be removed; young growth must be safeguarded against injury.

79. Markets must be canvassed in order to develop a sale for the lesser species which crowd the forest and stand in the way of profitable regeneration and improvement.

80. The areas most in need of cutting should be located accurately and operations diverted thereto.

81. Present timber-getting operations should be limited, by the prescription of a high minimum girth, to the skimming from the forest of the overmature and deteriorating trees.

82. Secondary fellings should be in the nature of a seeding cutting, and should be conducted, when practicable, by the forest service itself. Otherwise, the forest service should mark the trees to be removed.

83. Marking rules should be developed in the light of the silvicultural lessons learnt. The work of marking should be regarded as of the first importance, and be under the direct superintendence of expert officers, who should establish sample markings on each cutting area.

84. A systematic study of cut-over areas should be maintained, and the results recorded at the Forest Institute for use in developing silvicultural practice.

85. The processes of natural regeneration should be observed closely in the forest, and worked out on experimental areas and in forest nurseries. The time of seedfall should be recorded for each species. Sample plots of seedlings should be marked for observation. The influences of firing, light regulation, shelter, soil-working, &c.-and their costs-should be analysed.

86. Forest improvement in the sense of silvicultural tending should be carried on, for the present, only experimentally, and as such should be studied closely as a part of research.

Undoubtedly, it is desirable that areas already cut over under the unrestricted conditions of the past should be set in order. But it is a tremendous and costly task, and general forest organisation should be regarded as of prior urgency, more especially as sound silvicultural practice has yet to be developed, and to a large extent also is dependent upon a widening of the Australian market.

87. A demonstration area should be set apart on each forest for direct exploitation and regeneration by the Forest Service, to serve both as an example of conservative utilisation and an experiment in applied silviculture.

88. In opening up roads within the forest, the Forest Service should itself convert and dispose of the trees to be removed.

89. Thinning work should be undertaken only where a market for thinnings exists or can be established. As a rule, there is little need for it in Australian forest practice, unless in the way of improvement fellings in the overwood. Congested pole crops are of infrequent occurrence.

90. A maximum periodic cut approximating to the sustained yield should be fixed for each forest.

91. Increment plots for the purpose of recording the rate of growth for different species, and accurately determining the rotation, should be established on each forest and for each site quality.

92. As the forest approaches normality, it will be possible to increase the periodic cut. The present annual sustained yield for the Brooloo Forest (Q.) of 40,000 acres has been fixed at 4,000,000 feet. When fully stocked, it is estimated that it will produce from 20 to 40 million superficial feet of hoop pine per annum.

93. Pending their organisation, State forests should be withheld from immediate exploitation, whenever it is practicable to divert cutting to timber reserves, ordinary Crown lands, leaseholds, or private property. I regard this policy as of paramount importance to Australian forestry. The timber resources carried by those lands must be utilised to the uttermost. Every stick destroyed in clearing for settlement means equivalent loss to the permanent reservations.

94. A silvicultural advance may be initiated by the attachment to timber sale agreements or licenses of essential provisos. Compensation for any added labour cost should be arranged in fixing royalties.

95. The correct appraisal of stumpage values is fundamental to the institution of forest management. An intensive study of labour, freight, and

market conditions is essential to its solution.