out such places over a big area? And did you ever plot and compare curves showing the area burned over per week throughout the fire season, the number of fires per week of the season, and the number of men per week of the season necessary to give the proper protection, and at the same time keep costs to a practical minimum ?

Fire protection is undoubtedly a big and vital question, so vital that more time should be given to it in schools and colleges than has been done in the past. It is true that fire fighting cannot be taught. But fire fighting is secondary to fire protection and fire prevention and to a large extent these can be taught.

In any course on fire protection, prevention should also be considered. Fire prevention is largely a question of publicity, organization, personality and legislation.

No matter how good the local or national legislation is, it may be practically nullified unless the public is brought into sympathy with the objects sought. This can best be done with local residents by personal contact, and with other users of a forest area by publicity. We should all recognize the fact that we must get the public interested in our work. To do this, interesting and instructive articles and so called news items must be prepared. The importance of this phase of the work should not be overlooked.

In fire protection much preliminary work is necessary. Fortunately the members of the Harvard (and several other) schools should have, by now, much of the preliminary data.

There is required, first, as intimate a knowledge as is possible of the area to be protected, its topography, climate, prevailing winds, moisture conditions, forest conditions (including brush, debris, especially valuable sub areas, etc.), its existing roads, trails, railroads, telephone lines, towns, lumber, and other camps and ranches; the principal area used by campers, hunters, and fishermen, and the main routes used by them in entering and leaving the area under consideration.

In addition there should be obtained a history of recent fires, their location, area, causes, damage done, and methods and cost of fighting them. This information should be tabulated and summarized. As much of this information as is possible should be shown on a topographic base map of the area.

Right here should be emphasized the necessity, (first) for careful and detailed plans and, (second) the putting of those plans on paper. The man who doesn't plan this part of his work carefully and thoroughly ought to be given a job requiring a strong back and a weak mind, rather than vice versa. And the man who, knowing that he personally can't see to all the details of every emergency that may arise, doesn't, after making them, put his plans on paper, should be treated only a little more leniently than the first man.

All this, and much more information is but preliminary to the actual plan. Next should come the determination of the number of men needed to insure protection, the points at which they should be placed and the lengths of time they should be employed. Will a system of stationary lookouts on high and commanding peaks serve, or do you need mounted patrolmen? Should you have a combination of the two or do you need moving lookouts? Should the patrolmen stick to the ridges in order to see over a big country or should they be in the main-used canyons and on the maintravelled roads and trails to keep track of hunters, campers, and fishermen? Have you got commanding peaks or ridges that can be used or trails that will allow patrols to travel in the right places ? Shall the patrols go on foot, horseback, or should they use motorcycles or automobiles? When does the fire season begin does it end and what is the most dangerous season? will give you the best protection, ten men from the beginning to the end of the season (say five months) at a total cost of $3750.00 or one man for five months, two for four months, six for three months, nine for two months, and one for one month. In the second case it is quite probable that you are getting better protection with the same amount of money by employing twenty men for short periods instead of ten men for longer periods. On the other hand, it may be possible that conditions are such that a number of comparatively long term men must be employed.

when Which

If lookouts are used there must be means of quick transmission of messages, either telephone, telegraph, heliographs or a wig wag system. The patrols should be able to get messages to headquarters quickly and they must be able either to get men and supplies themselves or have them sent in without delay. What methods of communication are available and what more are

essential?

When should they be built and will their cost pay in the long run? Have you made arrangements for the lookouts to live on top of the peaks or will they camp below somewhere where they will lose two hours through the day and all the evening, night, and early morning?

It is essential that your patrols keep as much as possible in touch with headquarters, and you or some one, should know where to get them at any time of the day or night. Are they travelling their particular country promiscuously or have you laid out for them definite routes, with schedules, so you will know where to intercept them if necessary?

Does each man of your protective force (outside of the stationary lookouts) know where he can get men, tools and supplies quickest for any fire that may occur? How is he to provide transportation? Is there a store at which provisions can be obtained and does he know how many provisions to take without packing an extra 100 or 200 pounds? How many extra men can he get, where can he get them, and how soon can they reach the fire, if it is a big one and he needs a big force?

These are a few of the questions which nearly every one of you will have to consider. The answers will be different in nearly every case, but if you will consider the problem for a given area now, it will help immensely when you start out into actual work of your own.

NOTES ON THE CHESTNUT BARK DISEASE
(DIAPORTHE PARASITICA, MURRILL) IN
PETERSHAM, MASS.

J. KITTREDGE, Jr., 1913

THE following data were collected in October and November, 1912, at Petersham, Massachusetts. The object of the study was to determine the present status of the chestnut bark disease in and about the Harvard Forest and to secure any possible information about the habits and appearance of the fungus which might indicate the means by which it is or is not distributed.

The disease was first noticed in Petersham in the fall of 1910 by Professor J. G. Jack of the Harvard Forest School, along a road not over a mile from the School and about three and one-half miles north of the village of Petersham. In August and September, 1911, Professor A. H. Graves, then in the Bureau of Plant Industry, found infection at several places in and about Petersham, particularly along the Athol Road. At that time he marked infected trees in several localities and carefully examined others near them, noting whether they were probably infected or apparently sound. The locations of a number of these infectious were pointed out to the writer by Professor E. E. Carter of the Forest School.

In this study the infection was viewed from the forester's standpoint, with little or no reference to the morphological or physiological characteristics of the disease. At the time of the study, only the winter fruiting bodies were in evidence.

In the first part of the study, data was collected as follows: All trees showing any evidence of infection were first classified in three groups:

I. Trees surely infected (with the fruiting bodies developed). II. Trees probably infected (with dead branches and persistent dead leaves or burrs or badly sunken bark).

III. Suspicious trees (with the persistent leaves or sunken bark less markedly developed than for Class II trees).

Trees which were apparently uninfected were entirely disreFor all trees included in the above tree classes, the following data was noted:—

garded.

1. Careful geographical location.

2. Origin, seedling or sprout.

3. Diameter at breast height.

4. Crown class; whether dominant, co-dominant, intermediate, overtopped, or suppressed.

5. Distribution of infection about known centres, and relations of crown and distance between infected specimens (numbered and oriented sketches).

6. Location of trees relative to site, aspect, forest type, and density of surrounding stand or degree of isolation with history of its cause.

7. Probable length of time since infection started.

The field work consisted in the inspection of trees, along roads, on strips through the woods, and around infections which had been previously reported. After some data had been secured in this way, the rest of the time was devoted to a detailed study of one stand on the Harvard Forest, which will be described later.

The infection is now almost universally distributed in the town of Petersham. Surely infected trees were found from the Athol line on the north to North Dana and New Salem on the west, and two miles south of the village toward Barre. The region to the east of the village was not examined.

Origin seems to have had no effect on the susceptibility of chestnut to the disease. The data obtained showed that the total number of infected sprouts in Classes I and II exceeded the total number of infected seedlings in the ratio of two and one-half to one, but this difference is easily accounted for by the predominance of sprouts over seedlings in the whole chestnut growth of the region. When all trees are included as a basis of comparison, the ratio is about three and one-half to one. This is, however, in a stand typically of sprout origin. On the special plot, the insignificance of the influence of origin is more evident, for of 219 trees of seedling origin, 32 (14.5 per cent) were surely infected, and of 747 trees of sprout origin, 97 (13 per cent) were surely infected. Trees in Classes II and III have been omitted from consideration