ASENTS FOE THE SALE OF MADRAS GOVERNMENT PUBLICATIONS.

IN INDIA.

R. Cambbay & Co., Calcutta.

Combeidgb & Co., Madras.

T. Coopoosawmy Naikeb & Co., Madras.

Higginbotham & Co., Mount Road, Madras.

V. Kaltanaeama Iyeb & Co., Esplanale, Madras.

G. C. Loganadham Bbotheks, Madras.

S. MuETnT & Co., Kapalee Press, Madras.

G. A. Natesan & Co., Madras.

P. R. Rama Itae & Co., Madras.

Radhabai Atmaeam Sagun, Bombay.

Sabasvati Publishing House, Popham's Broadway, Madras.

• T. K. Sitabama Aitae, Kumbakonam.

D. B. Tabapobevala Sons & Co., Bombay.

Temple & Co., Georgetown, Madras.

Tiiackeb & Co. (Limited), Bombay.

TnACKEB, Spink & Co., Calcutta.

Thompson & Co., Madras.

IN ENGLAND.

B. H. Blaokwell, 50 and 51, Broad Street, Oxford,

Constable & Co., 10, Orange Street, Leicester Squaie, London, W.C.

Deighton, Bell & Co., Cambridge.

T. Fisiieb Unwin, 1, Adelphi Teriace, London, W.C.

Gbindlat & Co., 54, Parliament Street, London, S.W.

Kegan Paul, Teench, Tbubnee H Co., 43, Gerrard Street, Soho,

London, W.
Henbt S. King & Co., 65, Cornhill, London, E.C.

P. S. King & Son, 2 and 4, Great Smith Street, Westminster, Loudon, S.W.
Luzac & Co.. 46, Great Russell Street, London, W.C.
B. Quaeitch, 11, Grafton Street, Now Bond Street, London, W.
\V. Tiiackeb & Co., 2, Creed Lane, London, E.C.

ON THE CONTINENT.

Feiedlandeb & Soiin, 11, Carlstrasse, Berlin.
Otto Habbassowitz, Leipzig.
Kabl W. Hiebsemann, Leipzig.
Kbnest Leboux, 28, Kne Bonaparte, Paris,
Mabtihts Nijhoei, The Ha^ue, Holland.

Agent for sale of tbe Legislative Department publications only.

—

d

H
W

t— i

S

C/3

t/2

MADRAS FISHERIES BUREAU.

MARINE FISH-FARMING

FOR INDIA.

BY

JAMES HORNELL, f.l.s.,

Ma line Assistant. Madras Fisheries Bureau.

VOLUME 1 1.— BULLETIN No. 6.

MADRAS:
PRINTED BY THE SUPERINTENDENT, GOVERNMENT PRESS.

[Price, i rupee 4 annas.] 191 1. [2 shillings.']

1L ^0

■i

CONTENTS.

Page

Introductory i

French Fish-farming at Arcachon 4

Construction of sluices ... ... . 7

Methods of stocking with fry . ... ... ... ... ... 9

Freshening the ponds ... .. ... ... ... ... ... n

Feeding and care of the fish ... ... ... ... ... ... 12

Wintering ponds ... ... ... .. ... .. ... .. 13

Methods of capturing the fish reared in the ponds 14

Age and growth of mullet, eels and bass ... ... ... ... 16

Extent and profits of the Arcachon industry ... ... ... 16

Descriptions of the photographs reproduced ... ... ,.. 19

The Communal Fish-Farms of Comacchio 21

Descriptive and historical ... ... ... ... ... 24

The stocking of the lagoon 31

The construction and use of tresse 33

Methods of fishing the lagoon 36

Description of a fishing labyrinth ... ... 39

Boat-locks in Service-passages ... ... ... 41

The causes of bad seasons ... ... .. ... ... ... 41

Remedies suggested ... ... ... ... ... ... 46

The disposal of the products 47

Method of marinating eels ... ... ... ... ... ... 50

Do. do. smelts ... ... ... ... 51

The storage and transport of live eels 52

Administration .. -.. .. 53

Revenue 55

Proposed reclamation of the lagoon ... . ... 58

Conclusions 59

The Scope for marine fish-farming in India ... 63

The species of fishes suitable 63

Table of salinity tolerance of certain fishes ... ... ... 65

Life-cycle and migrations of eels ... ... ... ... 68

Habits and growth of eels ... ... ... . ... ... 70

Habits and growth of mullet ... ... ... ... 72

Habits of Polynemids 73

Habits of Pomfret 74

Do. Sea-breams 75

Do. the Koduva (Lates calcarifa) ... .. ... ... 75

Do. Sillago sikama, the Madras " whiting " ... ... ... 75

The spawning time of smelts 76

Flatfishes ... ... .. ... ... 76

Catfishes 76

The Waters available and suitable in the Madras Presidency. 77

Backwaters and estuarine creeks 77

Salt-pan channels 79

Estuarine and deltaic marshes 80

Procedure suggested to meet local conditions in India ... 81

.*

EXPLANATION OF PLATES.

PLATE I (FRONTISPIECE).

General view of San Carlo Fishing Station and labyrinth, Valle Ponti, Comacchio.

In the centre is seen the terminal otele ; on its massive reed palisade-wall a fisher-
man is standing with the dip-net in his hand used for lifting out the trapped
eels. To the right is a stack of bundles of reeds employed in the construction
of the labyrinth walls. The building behind is used as barracks by the fishermen
stationed here.

PLATE II.

Figure i. Operation of freshening the water (faite boite) of a fish-farm at Le

Teiche, Arcachon. View of the inner end of a wooden sluice.
One sluice shutter is raised to its full extent and the tide in high flood is pouring

through into the ponds. Immediately under the raised wire-net screen is seen

the distended bag of the sleeve net.
Figure 2. Seaward end of the same sluice taken at the same time as figure 1.

These two figures show well the simple construction of the majority of sluices

in use.

PLATE III.

Figure 3. View of the seaward end of the same sluice as in figures I and 2, during
the operation of deboire — the letting out of a portion of the pond water prepa-
ratory to " freshening." In this operation the photograph shows that the
sluice shutter is raised very slightly in order to obviate the production of a
strong current from the ponds to the sea.

Figure 4. The same sluice closed. Note the great length of the sleeve-net sus-
pended between its frame and the sluice-gate for the purpose of drying.

PLATE IV.

Figure 5. Le Teiche fish-farm. Seaward end of a closed sluice having concrete
walls. The outer wire-net screen is lowered to rest in its grooves. The flat
nature of the adjacent country is clearly indicated.

Figure 6. View of the inner end of a double sluice-way at Tacquets, Arcachon.
The sluice-gates are closed and the sleeve-nets raised for drying. Note the
solid construction of the sluice walls and the substitution of iron shutters and
screws for the wooden ones seen in the preceding illustrations.

PLATE V.

Figure 7. Fish labyrinth, Serilla Fishing Station, Valle Vacca, Comacchio.
View from the covola looking towards the apex of the first set of palisades.

Figure 8. The same labyrinth to show a lateral otele wherein the eels are finally
trapped. The group of men at the left of the view are standing about the
baldresca, the thin-walled chamber where the mullet are separated from the
eels. A moveable hooped fish-trap is seen suspended from a pole over the
otele.

PLATE VI.

Figure 9. View of the terminal trap-chambers (oteli) of two twin labyrinths at
Serilla Fishing Station. The substantial nature of the piles and crossbeams
supporting the reed palisades is well seen.

Figure 10. View from within the cogolara of one of the above labyrinths looking
towards the mouth of the terminal otele or trap. Note the great thickness of
the reed palisading and its inner and outer series of supporting piles.

PLATE VII.

Figure ii. View from Serilla Fishing Station looking towards its sea-canal (Canal
Ungola). Note the nets stretched across the mouth of the canal to prevent
escape to the sea of any fish which succeed in forcing the passage of the
labyrinths.

Figure 12. A lock in a service passage leading from a sea-canal into Valle Cona.
See text-figure No. 8 for the outline of a lock-shutter.

b

MARINE FISH-FARMING:

A DESCRIPTION OF

FRENCH AND ITALIAN METHODS

WORTHY OF TRIAL IN INDIAN

BACKWATERS AND DELTAS.

INTRODUCTORY.

The utilisation of ponds and lagoons for the rearing
of fish from fry is an industry of great antiquity. The
Egyptians appear to have elaborated suitable methods
2,500 years ago and we know that the Chinese have
been successfully engaged in similar pursuits for an
unknown number of centuries.

In Europe fish-farming" was first practised by the
Romans. We know that the fattening of sea-fish was a
fashionable pursuit of the luxurious during the first
century B.C. Pliny and Terentius Varro have given
descriptions of these fish-ponds, amongst the most famed
being those formed by Lucullus near Naples which com-
municated with the sea by a canal reputed to have cost
more than the amount lavished by the noble owner upon
the construction of his villa. This canal was fitted with
sluice gates and by methods very similar to those to be
described in detail on a later page, both the fry and the
adults of several species of sea fish were induced to
ascend the canal to the ponds. As showing the great
scale on which these ponds were conceived and the suc-
cess of these early operations, we have Pliny's statement
that after the death of Lucullus, the fish in his ponds
realized four million sesterces or not less than £40,000.

The Roman marine fish-farms in spite of their great
cost are reputed to have yielded handsome profits when
experience in their management was gained. None
appear to have survived the Augustine age though
it may well be that the establishment of fish-culture in

Comacchio lagoon owed its inception to men conversant
with Roman methods of fish-farming.

Down to within a hundred years of our own time,
the Comacchio and other Italian establishments on the
Adriatic had no imitators and to-day marine fish-farming
worthy of the name is limited to Italy and France. In
view of the great attention devoted to the culture of
fresh-water fishes during the past 60 years and the
wonderfully good commercial results obtained therefrom
especially in the United States of America and in Ger-
many, this neglect of the pond culture of sea fish may
appear strange. Two principal reasons may be given —
the higher capital expenditure usually necessary to con-
struct and organize marine fish-farms owing to the larger
scale on which they have to be conceived and the general
ignorance of the lines on which such ponds must be
constructed and regulated. We have innumerable works
in English upon fish-culture, but the authors with sin-
gular unanimity limit their descriptions and instructions
to fresh water work ; indeed the majority of pisciculturists
appear to be unaware of the fact that sea-fish culture in
ponds is practised with profit in France and Italy.

The most suitable coastal districts for marine fish-
culture are those where there is a wide margin of shore
lands scarcely if at all raised beyond the level of high-tide.
Such lands have been formed in the main as deltaic de-
posits at and around the mouths of rivers. Others again
arise from the drifting of sand and other debris of the
land along a coast by the force of locally prevailing winds
and currents and their accumulation as fiats and dunes in
angles of the coast line. Prevailing winds and currents
may also form long sand spits parallel with the coast at
the mouths of rivers and, by deflecting the river estuary
in the same direction, create estuarine sheets of water
often of very great area ; these are known as backwaters
in India, and as lagoons in Italy.

Deltaic marshes and shallow backwaters provide the
best conditions for sea-fish culture on commercial lines
and both are numerous on Indian coasts. The Madras
coast line is specially favoured ; along the western coast
stretches a long chain of backwaters frequently connected
and giving 0ff innumerable branches and side creeks ;
the eastern coast also oossesses a number of important

backwaters, as Covelong, Cuddalore, Ennore, Pulicat,
and in addition possesses a great number of extensive
river deltas — enormous areas of low marshy land inter-
sected by a network of channels.

With a view to study the details of fish-culture suit-
able for Indian conditions, the adoption of which may
make it possible to utilize eventually some considerable
area of the Madras backwaters and estuarine marshes, I
took the opportunity last year (1909) when on leave in
Europe, to visit the two localities, Arcachon in France,
and Comacchio in Italy, where the rearing of sea fish in
ponds and lagoons is carried on very extensively accord-
ing to practical methods which have stood the economic
test of commercial conditions during several centuries in
the case of Comacchio and over 60 years in that of
Arcachon.

I-A

I.— FRENCH FISH-FARMING AT
ARCACHON.

The physical features of the basin of Arcachon where
are situated the most famous of French marine fish-
farms were described briefly in Bulletin No. 5, pp. 4 to 7.
As there stated the basin is a great backwater roughly
triangular in outline. Its area at full tide is about 35,000
acres, or approximately one-third the size of Pulicat Lake
near Madras. Like the latter Arcachon basin is mar-
gined along many sections of its shore by low-lying
marshes. These are specially extensive towards the
inner or eastern end of the basin where the river Leyre
debouches through a network of channels. These del-
taic flats and marshes when not protected by embank-
ments are liable to be overflowed during spring tides,
and at a very early date a large acreage was converted
into salt pans by the erection of dams fitted with sluice
Sfates. With the sea water admitted from the basin from
time to time to supply tue pans came abundant fry of
eels, mullet, and bass ; these found refuge in the connec-
ting ditches and in occasional deep pools where the salt
workers saw them thrive and fatten and found them a
substantial addition to their meagre fare. From this to
intentional reservation of definite areas for fish-rearino- is
a short step, especially as it must early have been noted
that such fish -rearing is a less hazardous industry than
salt gathering where the harvest is at the mercy of a
fickle climate. Till the middle of the 19th century the
methods followed remained extremely primitive ; at that
epoch consequent upon improvement in the means of
transport and distribution due to the introduction of
railways, this industry received much attention, methods
were refined, and further areas devoted to the purpose.
The trade was a very profitable one for the proprietors
till about the end of last century, the expenses of upkeep
being very small, and the prisoned fish bringing excellent
prices in Bordeaux whenever stormy weather interfered
with sea fishing. To-day the owners have to face les-
sened profits owing to the erection of large refrigerating
stores at Arcachon and Bordeaux where reserve fish is
kept in stock to meet the emergency of bad fishing
weather at sea. In spite of this the fish farms are con-

sidered to give better returns than would be the case if
they were drained and put to agricultural or grazing
purposes.

These fish farms, or reservoirs a poissons as they are
named at Arcachon, consist of large shallow ponds and
broad waterways occupying the site of former salt pans
and marshes. There is no regularity or settled usage as
to shape or area ; the configuration of the old marshes
and the size of the proprietor's farm are the determining
factors. The principal ponds in each farm are however
as broad as possible and a usual size is from 300 to 600
yards long by 80 to 120 yards in width, but as I have said
there is no rule and some extend to twice and even three
times a greater area. Several ponds are found on each
farm connected by canals and separated by stretches of
low land which furnish scanty pasturage for cattle.
Each fish farm is separated from the basin by a protec-
tive dyke or bund of height and strength sufficient to
prevent flooding during high tides and storms. One or
more sluices in this embankment provide the means of
establishing connection at will between the ponds within
and the sea water of the basin without. These sluices
and their proper management are the pivots upon which
the system depends for its success. Through the
careful manipulation of the sluice gates at the right
season, outflowing currents are set up which attract the
shoals of fry required annually to repopulate the ponds.
Another important duty devolving upon the sluices is
to keep the water of the ponds in a wholesome and well
aerated condition during the period the imprisoned
fry are growing to the adult stage. This process of
freshening consists of two operations called respectively
faire boire and faire ddboire. The former allows the
entry of the well oxygenated sea-water of the basin,
charged too with quantities of minute life welcome to the
smaller creatures within the ponds — the younger fry as
well as the minute animals on which the fry feed. By
the converse operation, faire dcboire, is expelled a
proportionate amount of pond water more or less con-
taminated with decomposition products arising from the
dejecta of the fish and from decaying vegetation; this
water is also less well charged with oxygen than the
compensation water admitted from the sea.

My acquaintance with the working details of Arca-
chon fish-farms is due in the main to the courtesy of
Monsieur Garnung of Le Teiche, who kindly arranged
that I should have every facility to visit and inspect his
extensive fish-farm situated in the middle of the delta of
the Leyre. With true French versatility M. Garnung
besides owning this fish-culture establishment, owns and
superintends a large resin refinery, a well-equipped steam
saw-mill, and the electric light plant which furnishes
current for the street illumination of the neiohbourino-
town of Le Teiche ; private interests however do not
absorb all his attention for he still finds time to look
after the welfare of the town in the capacity of Mayor.

Twice I visited M. Garnung's fish- farm and on both
occasions I had the advantage of his personal guidance
and explanations. The tarm lies just below the level of
high-tides, between creeks which lead down to the
eastern boundary of the Arcachon basin. The seaward
side of the farm is protected against the tide by a strong
embankment formed of the clayey alluvium composing
the surface of the delta. Its height, governed by the
local vertical range of the tide, is usually about 12 feet ;
the breadth at the base averages about 36 feet, de-
creasing at the summit to 6 feet. The slopes are well
consolidated and are further strengthened by the growth
of grass and sedges. On the majority offish-farms the
area protected by the sea-dyke is divided almost equally
between land and water. The largest farm at Arcachon
(Audenge) comprises an area of 168 hectares (415 acres)
under water and 172 hectares (425 acres) of grazing land.
M. Garnung's farm is much smaller but the same ratio
persists. The water area is greatly split up, forming a
series of irregular ponds connected by wide channels.
Sluice gates divide the various ponds and these differ
considerably in depth and in arrangement. The shallow-
est ponds adjoin the sluices opening to the sea, the
deeper ones lie landward of the former. The seaward
or entrance ponds as we may term them, are wide shallow
expanses of water ranging according to the size of the
farm from quite small areas up to as much as 300 to 600
yards long by 80 to 130 yards wide. The depth usually
does not exceed 18 inches and frequently is no more
than a foot. Towards the margins the depth is greater,

ditch-like channels having been dug to furnish depths
carrying from 3 feet to 4J feet of water where the young
fish may retreat for safety upon any sudden atmospheric
change. In the shallows of the entrance ponds flourishes
a luxuriant growth of marine and brackish water vegeta-
tion, chiefly confervae and a species of marine flowering
plant {Ruppia maritima) which finds here extremely
congenial conditions. The marine prairie thus formed
over the shallow central portion of each pond is the
home of swarms of minute life — infusorians, zoophytes,
and small crustaceans and molluscs, constituting the main
food supply of the young fishes. The great majority of
these fish-farms and all the large ones have separate
wintering ponds where the fish are herded during cold
weather. These are deep canal-like ponds of sinuous
course, sometimes over a mile in length by 20 to 30 feet
in width. As a long length is required the course
doubles upon itself so often that it becomes a veritable
maze. These wintering channels are situated in the
most sheltered section of each property and care is taken
when excavating them to pile the material dug out as a
protective embankment along the north-east and south-
east boundaries as a protection against the cold winds
which blow from these directions ; as a further protec-
tion these banks are often planted with thick hedges of
tamarisk, a plant that thrives well on the coasts of
Western Europe.

Construction of Sluices.

The general plan and arrangement of the sluices
which place the ponds in communication with Arcachon
basin vary very slightly in the different farms ; the
dimensions which follow are those of a sluice on
M. Garnung's property and they and the accompanying
plans (text-figs. 1 to 3) and photographs (figures 1 to 5)
of sluices on the same farm may be considered as typical
of the form in accepted use. The differences which
exist are chiefly in the materials employed and in the
mechanical details of the means employed to raise and
lower the sluice shutters.

The sluice now to be described is one leading into
the principal entrance pond. It consists of a cutting

8

through the sea-dyke 32 feet long and 47 inches wide.
In the more modern farms the sides and floor are of
concrete, but in the older forms (figs. 1 to 4) the framework
is built up entirely of fir planking supported by piles.
Twenty-one feet from the seaward extremity two narrow
sluice shutters are fitted sliding up and down in a grooved
framework and actuated either by a screw adjustment or
by the leverage of a crowbar. From this point the floor
of the sluice slopes gently downwards towards each
extremity so that its level at the seaward end is
depressed 4 inches and that at the inner end 2 inches
below the level at the sluice gates. The floor of the
sluice is arranged so as to be from 3 to 4 feet below the
mean level of the entrance ponds and about 2 feet above
the bottom of the wintering ponds. Each extremity of
the sluice (S. and S.S. text fig. 1) is furnished with a

32 FT. 6 IN

l-Jttl m

1—6. N.

!
\k2 FTfc

tZ

MI

a s- 1-

Q

0

CONCRETE WALL

21 FT-

S. G.

I.

->

ffiU

•>

s. s.

CONCRETE WALL

"Li""1

Fig. 1. — Plan of sluice to show general design and dimensions.

S.G. Sluice gate provided with two shutters ;
S.S. Seaward wire net screen ; S. Inner screen.
S.N. Indicates position of frame of sleeve-net.

moveable wooden frame sliding in vertical grooves cut
in the side walls. Over each frame is nailed a screen of
strong wire netting of small mesh so that when in place in
its grooves it furnishes an effectual barrier to the passage
of all but very small objects. The overall dimensions of
these frames are 6 feet 9 inches by 4 feet 1 inch. Eight
feet behind the sluice gates and 2 feet in front of the
screen at the inner end of the sluice is another pair of
grooves accommodating a third frame. This instead of
being covered with wire netting as are the other two has
the mouth of a long tapering cotton net, the manche,

S5

o

<

o
a

o

C

c<->

C3

cS
ji

V

c >

U3

w

U-./3

o

s

H 5 S

55 «
.2 §

73 "~

2

■<

tut) •

.5 b£

O i/>
u cS

° Oh
>> UJ

4-* tA

"8 8

o JS

nailed or laced to its rectangular frame. This narrow
sleeve-shaped net has a length of 23 to 24 feet ; at the
mouth end the meshes measure 15-17 millimetres
(about f inch) and diminish to 10-11 millimetres (f inch)
towards the free extremity. Except when letting out
water from the ponds the frame of the sleeve net is kept
raised to the top of its grooves, the net itself being-
carried backwards and hung over the framework of
the sluice gates as shown in PI. Ill, Fig. 4. To
facilitate the working of the sluice a plank is laid
across behind each of the three movable frames (see text
Fig. 3.), and a plank way, 6 feet wide, crosses the
sluice at the summit of the embankment.

The Method of Stocking with Fry.

This operation depends upon the fact that the fry of
certain species of sea-fish resort to estuaries during the
early period of their life, led by an instinct which impels
them to swim against any current they encounter. Hence
to avoid the labour and expense of collecting the required
amount of fry in the sea by means of nets, the sluices of the
ponds are manipulated in such a manner as induces shoals
of fry to enter the ponds voluntarily. The operation is
conducted in the following manner. Each day during
the period of neap tides at the times when fry are expec-
ted, as soon as the tide begins to recede the wire screen
at the seaward end of the sluice is raised, the one at the
opposite end being kept in position; one of the sluice
shutters is then raised sufficiently to create a current
flowing from the pond to the sea. Following their instinct,
any small fish which are near the sluice swim against
this current and eventually find their way into the sluice
channel. When the man in charge of the operation
judges the time fit, either because a large number of fish
have congregated or because the tide has turned and
would shortly cause a change in the direction of the
sluice current, he replaces the seaward screen and closes
the sluice. The imprisoned fry are then fished out
of the sluice with the aid of a square-framed dip-net
and transferred to the adjacent pond where they are
kept till such as survive to maturity attain marketable
size and condition.

IO

Spring is the season when the bulk of the fry is
obtained. At this period grey mullet [Mugil labeo),
bass [Labrax lupus), and eels appear in great shoals ;
multitudes pass into the sluices and are captured. The
passage of the shoals of fry from the sea into estuaries
and even up rivers is technically termed montee by
French Culturists, a term which may be rendered into
English as the ascent of the fry. Although this takes
place principally in the spring in Western Europe, it
varies considerably with the species and scarcely any
month between February and October is without its
ascent. Thus at Arcachon the white mullet [Mugil
labeo) arrives as early as February and continues into
March. The common eel [Anguilla vulgaris) arrives
almost as soon in vast multitudes of strange transparent
larvae already about 7 centimetres (2 J inches) in length.
Bass appear chiefly during April and May, while the
shoals of the black mullet (Mugil chelo) do not usually
come till May and June. Another and greatly esteemed
mullet, Mugil cephalus, is still later in putting in an
appearance, September and October being its months of
ascent. The bulk of the fry other than of eels enters
with a size of from § to 2\ inches — the great shoals
are constituted of fry of this size. With the advance of the
season the young fishes attracted into the sluices de-
crease greatly in number concurrently with considerable
increase in individual size.

Of the two operations involved in the freshening of
the pond water, that involved in the running off of a
portion, the operation of faire deboirc, is conducted very
much in the same manner as the manipulation of the
sluice when shoals of fry are to be captured. As in the
latter operation, when the tide begins to fall, the screen
at the sea end is either removed or raised sufficiently to
give a clear passage, and then one or both of the sluice
doors are raised very slightly, usually from 3 to 6
centimetres (1^ to about i\ inches) and never
more than io centimetres (4 inches). By this means a
stream of water flows from the pond into the sea and
continues to flow not only till low tide, but also during
a portion of the flood and until the rising tide attains an
equal level with the water of the pond. At this moment,
unless it has been done before, the sluice doors

1 1

are raised and the seaward screen dropped home into its
grooves. A skilful operator can so judge and control
the strength of the outflowing current as to keep it
so weak that it has no power to carry out any of the fry
in the ponds — it is for this reason the sluice doors
are raised so slightly. The natural instinct of the fry to
swim aoainst a current is also counted on, while to
prevent larger fish passing out the screen at the pond
end of the sluice is kept lowered during deboire. Occa-
sionally however the fry do endeavour to pass out in
some quantity ; when this is observed the sluice shutters
are at once closed. When the operation is finished for
the day a considerable quantity of fry and prawns are
frequently found congregated in the sluice ; these are
caught and put into the ponds in the same manner as
when stocking the ponds with fry. It must be noted
that this operation of emptying a portion of the pond
water is carried out usually during spring tides, whereas
the operations for catching fry in quantity are carried
out during neaps. This and the greater height to
which the sluice shutters are raised durino- the latter
operation — involving a stronger current — are the two
important differences.

The converse operation of faire boirc, or actual
"freshening", is merely the reversal of the two opera-
tions already described. It consists of a regulated inflow
of sea water during spring tides equal to the amount
expelled during deboire and lost by evaporation. In
this operation both the wire-net screens are raised
(PI. II, Figs, i and 2) and then before opening one of
the sluice shutters the mouth frame of the sleeve-net is
dropped into position, the net itself being thrown into
the pond. As soon as the rising tide attains a higher
level than the surface of the ponds a sluice shutter is
raised slowly and with the utmost care in order that
the inrushing water may be adjusted with nicety to a
strength sufficient to distend the sleeve-net but not
violent enough to burst it. Text-figure 3 will make this
operation clear.

The scene in the pond during this process of freshen-
ing is extremely interesting (PI. II, Fig. 1). The great
24-foot long sleeve-net, distended and appearing taut to
the point of rupture, streams straight into the pond, and

12

all round its outer surface swarm hundred of dark bass
attracted by the current, all darting to and fro in the
greatest excitement either hunting for food particles
brought by the inrushing tide or endeavouring to find an
opening in the net and so pass up the current. Mullet
being shyer fish do not press forward so excitedly as the
bass do and one sees comparatively iew of them. A
baited hook dropped among the bass at this time is
seized at once and the veriest tyro may haul them out as
fast as the hook can be rebaited.

As in the other sluice operations, "freshening"
requires a nice judgment on the part of the gateman, a
judgment acquired only after years of experience. Thus
while it may be stated as usual for freshening, in its
two-fold operations of causing first an outflow and then
an inflow, to be carried on twice a day during spring-
tides from the middle of March till the beo-inning- of
November, there is no fixed rule ; the frequency and the
duration of each freshening are left to the discretion of
the man in charge who acts according to the varying
circumstances of weather, tide, and the condition of the
pond water. To a large extent the gate men seem to be
guided by that peculiar instinct or intuition gained by
fishermen from years of familiarity with their calling
which tells them unerringly that certain conditions are
present though they may be unable to explain how they
arrive at the conclusion. Usually the ponds are fresh-
ened at each tide for two days before and three days
after new and full moon. During the winter season,
November to March, it is not found necessary to carry on
these freshening operations, the entrance ponds being
then wholly or partially dry, and the fishes removed to
the wintering ponds.

Feeding and Care of the Fish.

Every time freshening takes place a certain amount
of food matter is passed into the ponds. In part this is
microscopical life which passes the screens freely, partly it
is small fish and shrimps caught in the sluice and thrown
into the ponds by the gateman. Beyond this no food
is given to the imprisoned fish, which are expected to
forage for themselves in the aquatic herbage flourishing

13

in the ponds. No use of artificial food is made, though
to any one conversant with the pond culture of
fresh- water fishes it appears certain that such a practice
would very greatly increase the success of this industry :
it is obvious that a very large proportion ol the fry ad-
mitted to the ponds is devoured by the stronger among
the bass and eels, and were a sufficiency of suitable arti-
ficial food added to the dietary of the fish, cannibalism
would be reduced, fatter and larger fish grown and a
greater quantity reared to maturity. It happens how-
ever that existing methods give fairly remunerative re-
sults and the fish-farmers are therefore satisfied to leave
methods as they are.

During their early days in the ponds and as long as
warm weather continues, the fry frequent the shallows
where they keep together in shoals, usually affecting the
sunny side. On the approach of winter the mullet and
the bass seek out retreats in the deeper parts of the
ponds, while the eels burrow into the mud. The mullet
espec'ally suffer severely from any considerable lowering
of the temperature and to protect these fishes in parti-
cular the deep wintering ponds already described are
constructed whenever the proprietor can afford them ;
in the case of very small farms where the stock of fish
held does not justify the expense ol such special provi-
sion, deep holes are dug here and there in the bottom of
the ponds and these are found to furnish a satisfactory
alternative when the protection of a limited number of
fish is in question.

On the farms where wintering ponds are provided
special operations of great delicacy have to be undertaken
in order to herd the fishes into their cold-weather
quarters. This is effected with the aid of water gates
which are interposed between the various sections
of the series of ponds. The procedure is similar to that
employed in the attraction and capture of fry for stock,
but is much more difficult to conduct owing to the
smaller volume of water dealt with. First the level of
water in the shallow summer ponds is reduced to the
utmost by running out as much water as possible through
the sea-sluices at low tide, the doors between the summer
and winter ponds being closed the while. When no
more water can be got rid of, the sea-sluices are shut.

H

The level of the deep winter pond is now at a higher
level than that of the summer pond, so when the doors
separating these are raised a sufficient head of water is
obtained to induce an appreciable current from the deep
winter pond into the shallow summer one. The most
experienced hand superintends the operation and he has
need of all his skill for the volume of water available
used to produce the current is very limited and does not
permit of a long-continued How. The flow must not be
permitted to be too great or the winter pond will empty
so rapidly that an equilibrium of level between it and
the summer pond will be established long before a
proper proportion of the fish have all passed upwards.
Neither must the current be too weak for then the
fish will fail to be attracted towards it. The operation
is repeated until all the fish have passed into the deep
wintering ponds. By the time this transfer is completed
the shallow summer ponds, now denuded of fish, contain
little or no water except in the deep ditches along their
margins and in this dry condition they remain the whole
winter. This annual draining and exposure of the bed of
the summer ponds to atmospheric influence is an opera-
tion of prime importance to the continued prosperity of
these fish -farms. An instance of the fatal result attendent
upon the long-continued non-observance of this axiom of
pond-culture is detailed on page 46. A severe winter is
an anxious season for Arcachon fish farmers in spite of
all the precautions that can be taken. They fear parti-
cularly the cold dry wind from the north-east for a tempest
from that direction may so reduce the temperature of the
ponds that the stock of mullet and bass may be deci-
mated and even destroyed in a single night. During
ordinary frosty weather when ice forms over the ponds
care is taken to break holes in it at numerous intervals
to allow some facilities for aeration.

Methods of capturing the Fish.

When they have attained marketable size the fish
are caught as demand requires by one of two methods
according to the species that may be ordered.
Mullet, bass and eels are the only kinds which mature in
considerable numbers in the ponds and while the two

i5

former are taken whenever required by means of nets,
this method of fishing is not suitable for the capture of
eels. To obtain the latter in quantity recourse is had to
the attractions of the sluice current at the season when
the reproductive instinct of these fish renders them
eager to gain the open sea and there seek the abyssmal
spawning depths where they were themselves hatched.
This desire develops in greater degree as autumn pro-
gresses and during September, October and November
the adult eels in the ponds are ready to respond to this
call of the sea. Accordingly on the occasion of any
convenient spring tide during the months named the
sluice gates are manipulated during one or more night
tides, so that a strong current of sea-water is directed
from Arcachon basin into the ponds. This is obtained
by raising one or both of the sluice shutters at the
moment when the tide has reached its highest level.
Concurrently the wire screen at the pond end of the
sluice is lifted, that at the seaward end remaining
down. The adult eels attracted by the current thus
produced come from all parts of the pond towards the
sluice in the endeavour to gain the sea. Baulked of
escape by the seaward screen they accumulate in the
sluice till the operator judges the number to be sufficient,
when both the sluice gates and the pond screen are shut
down. The eels may then be removed at leisure by
means of a dip net and transferred to storage boxes or
live cases moored in a deep part of a pond, where they
will remain till sent to market. It must be noted that
large catches are made only when the nights are dark
and stormy ; it is useless to attempt to attract the fish
into the sluices by this method on calm or clear nights
or during daylight, whence the necessity to keep large
stocks of eels on hand in the store boxes.

Mullet and bass are more easily taken when required ;
a trammel net set in the evening may generally be de-
pended to yield a considerable number of large indivi-
duals when it is taken up in the morning. Hence these
fish are taken only when orders arrive and when the
quantity netted is greater than is necessary, the surplus
are liberated as the trammel net has the great advantage
among nets of inflicting little or no injury upon the fish
taken in it.

i6

Age and Rate of Growth.

Mullet and bass take generally three years to attain
marketable size, while Gemzoe * has proved that eels do
not seek to pass to the sea till after \\ to 7 J years sojourn
in the case of males and an even longer period in the case
of females, the great range in the length of the sedentary
stage being determined by local conditions and circum-
stances.

Mullet, which enter the Arcachon ponds as tiny fry
i| to 2 inches long, are stated to weigh on an average
at the end of three years about half a kilogram or
about 18 ounces under normal conditions. But this
rate of growth is considerably surpassed in the better
situated ponds and when a series of warm and sunny
seasons occur consecutively. Under such circumstances
Millet f has claimed that 1,000 mullet fry reared in the
Arcachon pond will yield 1,000 kilos of fish within
three years.

The growth of the eel is considerably greater ; Coste
has estimated that these fish when reared in ponds
attain a weight of 2\ kilograms (5^ lbs.) within 4 to 5
years, but this again varies greatly with locality and the
relative amount of food supply available, and the rate
of growth must be worked out independently for
each separate district. Comparatively little exact re-
search has been devoted to this subject of the rate of
growth of sea-fishes ; precise information upon it is
urgently needed together with an analysis of the relative
value of the factors which favour and retard growth.

Extent and Profits of the Arcachon Industry.

The total area of the ponds devoted to fish rearing on
the borders of Arcachon basin has been estimated at not
less than 300 hectares or about 740 acres. A few large
properties account for the bulk of this, one farm alone
having more than half the total area, others again do not
amount to a full hectare.

* K. J. Gemzoe, Age and Rate of Growth of Eels. (Report of the Danish
Biological Station to the Board of Agriculture, xiv, 1906. Copenhagen, 1908.)
f La Production animal et vegetale. Paris, 1S67.

i7

The yield is very variable. In one year, these 740
acres are estimated to have yielded 68,060 kilos of fish,
while in the succeeding year, the yield rose to 97,660
kilos.*

This is equivalent to a mean yield of 92 kilos per
acre in one year and 132 kilos in the next. Another
estimatet of the yield is that one hectare of water, well
cared for. will give 150 kilos of mullet and 50 kilos of
eels in a good season; this work out at 81 kilos per
acre. This, however, I understand, is considered too low
for really well organized and up-to-date culture, and I
was assured that 120 kilos of fish per acre is the actual
average present-day yield of large fish farms when con-
ducted skilfully and with enterprise. A balance sheet
quoted by Millet (loc. cit.) gives the receipts and ex-
penditure for a farm comprising 100 hectares (247 acres)
of pond surface as follows : —

Receipts.

Francs.
Sale offish, 30,000 kilos at 1 franc per kilo. — 30,000

Expenditure.

Francs.
Foreman ... ... ... ... ... 460

3 Gatemen ... ... ... ... 1,050

Nets, boats, freight of fish, etc. ... 2,500

Upkeep of sluices and embankments

(average) ... ... ... ... 800

4,810

25,19°

This is equivalent to a net profit of, say, 252 francs
per hectare or 102 francs, say £<\, per acre.

This return may appear comparatively small in view of
the statement often made that an acre of water produces a
greater amount of food than the same area of land ; this
assertion is however most misleading when stated baldly
and without qualification for while some land is exceed-
ingly rich and capable of yielding prolific crops, other
tracts are virtually worthless, and so it is with the sea and
rivers ; there are areas where fish may congregate,

* Gobin, La Pisciculture en eaux salees, Paris, 1S91 .
t G. Roche, La Culture des mers, Paris, 1898.

iS

brought together in multitudes for the purpose of spawn-
ing or to feed on some food matter accumulated perhaps
in some indentation of the coast under the action of a
current eddy, and here the yield per acre of surface may
be prodigious. Elsewhere and even in the same places
at another season fishes may be all but wholly absent.
Inshore waters generally do appear to furnish larger and
more valuable food returns to man than average quality
of agricultural land, and the absolute productivity of the
sea per acre is probably greater than that of the land.
This latter is however not the question before us, which
is the profit to be drawn by man from the cultivation of
selected areas of shallow water.

To assess the value to a district of fish farming such
as is practised at Arcachon it is more just if we consider
what the same area would yield were it cultivated in any
other manner, rather than make a comparison with the
profit derivable from rich agricultural land. We must
also consider the relative amount of labour necessitated
in the cultivations compared. Considering the matter
from this standpoint we shall see this profit of £^ per
acre in a much more favourable light. The proprietor
of the farm we have mentioned was able to make a
profit of just ,£1,000 on the working of 243 acres of
what was originally a salt marsh and this with the aid
of a surprisingly small labour force — 4 men in all. What
form of land culture can show such a record for 4 men's
work ? Again, were this marsh land not utilised for fish
culture, its value for agricultural purposes would be
comparatively small for several years, because the land
being below the level of high tides and saturated with
salt, it would be a long time before it became sweet
enough to yield good crops.

It must further be borne in mind that the culture
system pursued is of the simplest description, so much
so that in the preceding pages I have avoided the use of
this term wherever possible, preferring the more elastic
one ' farming '. True culture implies breeding and this
the Arcachonnais have never attempted ; neither do they
supplement the food supply of the ponds except by
turning into them such fry and small crustaceans as are
taken fortuitously when freshening the water. Up to the
present time the attention of the owners has been

19

concentrated upon the perfecting of mechanical means for
the capture and safe keeping of fry hatched in the open
sea ; their system is perfect in respect of these opera-
tions but it fails to rear a sufficient weight of fish per
acre as it makes no provision to supplement the
limited food supply produced naturally within the ponds.
As a consequence of this limited food supply cannibalism
is prevalent and but a comparatively small number of fish
survive. No intensive culture is attempted and until it
is, Arcachon fish farming will remain an industry only
partially developed and will be unable to yield the further
profits which a vigorous policy of development on scienti-
fic principles would certainly yield.

At the same time it will be noted with satisfaction by
French culturists that the produce per hectare compares
favourably with the much vaunted results of German
carp culture, an industry now highly organized and
widely spread throughout that empire. Estimates of the
annual produce of carp reared in ponds as given by
German scientists, range from 65^ to 164 kilos per
hectare *, and it seems certain that the average cannot
be placed at more than 1 10 kilos, whereas at Arcachon
we have seen that the ascertained yield per hectare for
one year was 2 2 6-8 kilos and for the succeeding year
325*5 kilos, per hectare, or 276 kilos average for the
two seasons. The superior results given by the empirical
French methods over those directed by German science
show how responsive fish-culture is to the crudest of
treatment in the presence of favourable natural condi-
tions, temperature being one of the principal. In
Germany the rigors of winter are so severe that for
months the growth of fish suffers arrestment, whereas at
Arcachon the winters are notably mild and inclement
weather is of short duration.

The illustrations on plates II to IV represent graphi-
cally the construction and method of operating the sluices
of Arcachon fish ponds. Figs. 1 to 4 show sluices
constructed of wood p'anking, while those on plate IV
are two examples of more modern forms where concrete
is the material employed for the walls and floor. The
sluice depicted in fig. 6 is of the most modern type

* Brandt, Ueber d. Stettiner Haff, Wiss. Meeresunt. Kiel Kovim,, 1S96.

20

and leads into a fishpond constructed by M. Lesca at
Jaquets on the western shore of Arcachon basin ; here, as
may be easily discerned in the view, are two twin sluices
side by side, each with a single substantial shutter,
worked by an ordinary screw. In all the other views
reproduced two narrow twin shutters are fitted within a
single sluice way, a method inferior to the Jaquets plan
as the doubling of the sluices minimises the danger of
mishaps and permits of greater rapidity in the carrying-
out of the different sluice operations.

Figs, i and 2, plate II, show the operation of faire
boire, the inflow of seawater into the ponds. Fig. 1 is
of the inner or pond end of the sluice, fig. 2 being of the
opposite or seaward extremity. The former view shows
the inner screen and one of the sluice shutters lifted
clear of the channel of the sluice ; the distended mouth
end of the sleeve-net may also be discerned immediately
under the raised screen. Angling for bass attracted by
the inrushing current is going on briskly under M.
Garnuno-'s direction. Fief. 2 shows further details of
the operation of faire boire while fig. 3 shows a view of
the same end of the sluice during the converse operation
of deboire. Note in this photograph the low state of the
tide as compared with that shown in fig. 2 ; also that
in this operation the sluice shutter is raised very slightly.
The next views, figs. 4 and 5, show closed sluices. In
fig. 4 the frame of the sleeve-net is seen raised and the
net itself carried back and thrown over the upper frame
work of the sluice gates. This position of the sleeve-net
is shown again in the case of the Jaquets sluice, fig. 6,
where it will be noticed that the gradient of the embank-
ment is considerably easier than in the case of those seen
in the other views. This is due to the fact that the soil
at Jaquets is composed largely of sand and is therefore
less easily consolidated than the clayey alluvium at the
mouth of the Leyre.

MADRAS FISHERIES, BULLETIN No. 6.

Plate II.

Fig. i. — Freshening of a fish-farm, Le Teiche, Arcachon.
View of inner end of a wooden sluice.

Fig. 2. — The same showing the appearance of the seaward end
at high tide. Screen and one sluice shutter raised.

[ /. Homell, Photo.

MADRAS FISHERIES, BULLETIN No. 6.

Plate III.

Fig. 3. — View of seaward end ok a sea-sluice during the operation

OF DEBOIRK, Le TEICHE. LOW TIDE.

Fig. 4.— Le Teiche fish-farm. Inner portion of a closed sluice
showing inner screen raised and sleeve net drying.

[/. Hornell, Photo. \

MADRAS FISHERIES, BULLETIN No. 6.

Plate IV

Fig. 5. — Le Teiche fish-farm. Seawarl end of a closed sluice
with concrete walls. Outer screen lowered.

Fig. 6. — Double sluice-way at Jaquets, Arcachon. Sluice
closed and nets raised and drying.

[/. Horncll, Photo.]

21

II.— THE COMMUNAL FISH-FARMS OF

COMACCHIO.

Comacchio as the seat of the most extensive and most
highly organized system of marine fish-farming in exist-
ence was brought prominently to the world's notice for
the first time by the epoch-making account of a journey
made by Coste, the father of French aquiculture, during
the years 1853 and 1854 by instruction of Napoleon III.
It is interesting to read in the official directions issued
to Coste in 1852 that he was desired to explore the
coast of France and of the Italian Adriatic with "a
" view to determine under what conditions extensive
"experiment in the propagation and acclimatization of
" marine animals should be organized " ; at the period
in question the potentialities of the artificial cultiva-
tion of oysters and fish were occupying wide-spread
attention in France and the national optimism already
saw in imagination the numerous salt marshes and land-
locked bays along their coasts converted into fish-ponds
and oyster parks.

The results of Coste's investigations were published
by the French Government in 1B55 and in prefacing the
dry details of the present day condition and organization
of the Comacchio fish-rearing establishment, I cannot do
better than translate the picturesque and graphic descrip-
tion furnished by Coste who thus introduced the same
subject : —

" At the time when barbarians were driving the civilised peoples
(of the Roman Empire) before them the population of Comacchio,
like the founders of Venice, sought refuge in the bosom of an immense
marsh, which, since the time when it was thus occupied, has been
gradually transformed into a great establishment for the exploitation
of the sea, where its methods attract fry hatched in the Adriatic
and gather in a harvest of fish when these become adult, by processes
as well conceived as those pursued by agriculturists in the sowing
of their fields and the reaping of their crops."

" Less favoured than its neighbour Venice, and unable because of
the inferiority of its position to aspire either to commercial sovereignty
or to the rewards of conquest, Comacchio applied its genius to per-
fect an admirable system- of dykes built with the mud of its lakes, made
firm with fragments of the shells living in its waters, and intersected
by numerous sluices opening into well-arranged canals which while

22

giving access to the open sea and to the fresh water of the rivers
bounding two sides of the lagoon, allow control at will either of the
entire lagoon or of any one of its compartments, with as much facility
as one can operate a laboratory model ; gigantic labour until now
without renown, modestly accomplished by plain men, resigned to the
rough discipline of the sea, to the monotony of barrack life, to the
sacrifice of sleep during stormy nights when tempests torment the
lagoon and lash its surface into commotion — satisfied as the price of
so much labour with a modest wage and the ration of lish which a
paternal administration distributes to them daily."

" If we imagine a fleet under the autocratic rule of an admiral
charged with the oversight of all needs, to have anchored in mid-
ocean, condemned there to subsist upon the produce of their fishing,
communicating with the rest of the world solely for the purpose of
transhipping fish into vessels coming to trade for it, we have the image
of this colony whose fishing stations and barracks are scattered like
ships of a squadron among the isles within this immense lake."

Since Coste's historic visit, several well-known
French, Italian, and German scientists interested in fish-
culture, have studied on the spot the lessons to be drawn
from the practice of this great Italian fish-rearing enter-
prise, but I cannot learn that any British enquirer has
ever inspected the lagoon. References to it in English
works are both scanty and vague ; for this reason and
also because Coste's book, besides being difficult to
procure, requires considerable amplification and amend-
ment in the light of present day knowledge if to be
of practical service to fishery authorities in India, in
the following pages I propose to describe with such
detail as appears material, the existing condition and
organization of the Comacchio lagoon, the system by
which fish fry are introduced, reared and eventually
captured, the methods employed to prepare the produce
of the fishery for market, together with notes upon the
causes which from time to time during the past century
have affected adversely the prosperity of the under-
taking. The inherent defects of the system, and the
modifications in plan and working which appear to me
needful to introduce in order to remedy these drawbacks
and to permit of greatly enhanced results, will receive
attention and then by making comparison between the
methods practised respectively at Comacchio and at
Arcachon, I hope to be able, by utilizing the best
features of the two systems, to outline a practical
working; scheme for the establishment of fish-farms in

23

the back-waters, salt-works, and salt marshes of our
Indian littoral.

Comacchio lagoon will be found intensely interesting
alike by pisciculturists and biologists, by students of the
municipal or communal management of industries, and
by the non-expert who merely wish to quit the beaten
track and see a district where the din of the outside
world is unheard, where habits and customs are little
changed from what they were centuries ago. Such a
visit I commend to any one passing through Venice with
three or four clays to spare, but so out-of-way is Comac-
chio that I bad considerable trouble to find my way
there. Although not more than 50 miles down the
coast from Venice, the tourist agents scarcely knew it by
name. The best way to reach the town appears to be
to take train to Ferrara, a run of about 2\ hours from
Venice, and there to change to the steam tram that runs to
Ostellato, a village near the western end of the lagoon.
From this point, the voyager must trust to luck for a
conveyance to Comacchio unless he arrives there at
4-30 p.m. when a ramshackle mail-cart connects with,
the train. If Venice be left by an early morning train,
a couple of hours may be spared at Ferrara to visit the
cathedral and the mediaeval castle. Between Ferrara
and Ostellato the tramway runs along one side of the
main road, first skirting the Ferrara canal for some dis-
tance and then through a rich alluvial plain given over
to the cultivation of beetroot and hemp. No actual
vineyards are visible, but the rows of poplars skirting
the road and demarcating the fields are connected by
rich festoons of vines, heavily laden with grapes at the
time of my visit.

At Ostellato, I had choice between a ricketty mail-
cart and a nondescript closed vehicle with worn-out
springs and a harness composed principally of string.
This latter after considerable hesitation I chose and
having deposited my courier within, I mounted the box
with the driver, who beguiled the tedium of the journey
by reciting verses from Petrarch and an account of his
own adventures under Garibaldi during the war of
Italian independence.

It was after 7 o'clock when we arrived at Comacchio,
which to my surprise I found to be a town of several

24

thousand inhabitants, boasting a cathedral, several large
churches and a quaint three-way bridge at the intersec-
tion of four canals.

Owing to its isolation scarcely any one can be found
in the town capable of understanding either English or
French, hence unless the visitor be conversant with
Italian, it is necessary to engage the services of an
interpreter when leaving Venice. In order to obtain
permission to visit the fishing stations in the lagoon
application should be made at the office of the Commune,
where I am sure every assistance in obtaining a boat
and a guide will be willingly given. The Communal
officials will probably be able to arrange for a visit to
the eel-curing establishment of the lessees of the fishery,
who are at present Messrs. A. Cornia & Co. ; if not, then
a personal application to this firm will no doubt receive
as courteous response as did my own. Among the
points of interest which particularly deserve a visit, are
the fishery station of San Carlo, a little to the west of
the town, that of Cerilla in Valle Vacca, the port of
Magnavacca, and one of the river-sluices, say that of
San Alberto at the extreme south of the laQ-oon.

t>

Description of the Lagoon.

The lagoon of Comacchio is a vast shallow salt-water
lake, irregularly circular in outline, situated in the south-
ern section of the delta of the Po. Along a portion of the
northern side flows the river Volano, while the river
Reno similarly bounds the southern extremity. On the
west or landward aspect it is shut in by marshes and low-
lying agricultural land, while on the east it is separated
from the Adriatic by a long belt of sand dunes (techni-
cally termed lido in Italian) which extends from one
river mouth to the other with breadth varying from two
to three kilometres. The circumference of the lagoon
is about 125 kilometres {jl\ miles), whereof 9 are
bounded by the Volano, 12 by the Reno, 74 by the
semi-circular curve of the mainland on the western
aspect, and 30 by the Lido.

In Coste's time the water-area amounted to about
40,000 hectares (98,844 acres), divided among 40 com-
partments or basins of varying dimensions known as
valli (or campi in the Comacchio dialect) ; to-day by the

VALLE C

VALL£.A

Fig. 4. — Plan of a Fishing Station in Comacchio Lagoon.

B. Fishermen's barracks.

E. Embankments separating the valli.

K. Lock in service-canal.

L. Lavoriero or labyrinth (shown in red).

T. Quadrangle of a Tressa.

S. Store-house.

(The portions above water level are indicated by dotted stippling\

25

reclamation of three of the shallowest and least valuable
sections, there remain only 37 sub-lagoons or valli
aggregating about 36,500 hectares (90,200 acres).

To every valle is apportioned at least one fishing
station, where centre the oversight and working of the
valle thus served. Here are situated the lavorieri or
labyrinths employed to trap the mature fish, the barracks
where lives the fattore of the valle with the company of
fishermen and guards under his control, and the store-
houses where are kept the live-chests required during
the fishing season together with the materials needful
for the upkeep of the labyrinth palisades.

The fishing stations are located usually at places
where natural islets existed at the time when the respec-
tive embankments were constructed. To economize
labour when forming the valli every advantage appears
to have been taken of the occurrence of chains of sand
banks and of islets — remnants in many cases of a
sea-bar or lido. The most convenient islet in each
valle embankment was chosen as the site of a fishing
station and there were dug several short cross channels
to connect the confined water of the valle within, with the
ebb and flow of the sea-channel without. In the heart of
this plexus of short canals, which radiate fanlike towards
the water of the valle, lie the islets whereon the fisher-
men's barracks and the various store-houses are built, as
may be readily understood by reference to text-figure 4.
The principal canals are blocked against the passage of
boats by the several series of converging palisades forming
the labyrinths, so to overcome this difficulty, subsidiary
narrow canals (service canals) are provided to give
passage to boats between the valli which adjoin at these
stations and between them and the feeder sea-channel.
The service canals are provided with simple locks (Plate
VII, figure 12, and text-figures 4 and 5) easily manipu-
lated, whereby ready intercommunication is maintained
while safeguarding the fish in the valli from escape.

As mentioned above the various valli or, as we may
term them, "fish-farms" differ greatly in size, ranging
from the immense Valle Mezzano of 17,407 hectares
(43,016 acres) down to the tiny Valle Molino of 71
hectares (175J acres). About 20 of these great fish-
ponds may be reckoned to be of first-class rank in

26

regard to area and fish- rearing capability ; the remainder
are either so small or suffer from such disabilities as to
be of minor importance.

The various sections are bounded and divided from
each other partly by natural divisions consisting of
natural tongues of land and chains of islands and sand
banks, and for the rest by artificial earthwork dykes
connecting the islands and miniature peninsulas into
continuous embankments. Where necessary the foot of
the dyke on one or both sides is strengthened with rows
of stakes. The earthwork itself is the sandy mud from
the bottom of the lagoon ; there is generally a large ad-
mixture of broken shells in it and this is considered to
have a beneficial effect in the consolidation of the
rampart. The sedges and grasses which grow freely
over the slopes are further useful factors in the work of
consolidation.

The sand-bank peninsulas and chains of islands
nearly all run in a direction north and south, roughly
parallel with the present sea-coast. The best defined
chain divides the lagoon into two almost equal parts.
The western or inner half is again subdivided by a dyke
into two distinct and very large fish-rearing valli, the
Mezzano of 17,407 hectares and that of Fossadiporto of
2,701 hectares. The eastern or seaward half of the
lagoon contains the remainder of the fish-farms, very
numerous, very irregular in outline and often compara-
tively small. The depth of water in the valli is never
great, it ranges generally between 3 and 6 feet but in
times of drought large sections of the bottom become
drv through the shrinkage of the water area.

All the valli are in more or less direct communication
with the sea by means of canals opening from one of
two sea-mouths traversing the Lido, that natural break-
water which protects the lagoon from the inroads of the
sea. The principal sea-mouth, that called the Port of
Magnavacca, cuts through the Lido almost midway be-
tween the embouchure of the Volano and the Reno. Its
length is about 2 kilometres, by 40 metres (131 "2 feet)
wide, and 2J metres (8*2 feet) deep at mid-tide. The
banks are heavily timbered and two wooden groynes
or breakwaters defend the seaward end. This mouth
feeds the whole of the fish-farms with sea-water, with

27

the exception of a few in the north-eastern section of
the lagoon which are served from a small canal called
Bocca del Bianco, a natural sea-channel through the
Lido, scarcely 12 metres wide, not always active, and
serving directly two valli only.

The channel of Magnavacca after traversing the
Lido bifurcates ; each branch almost at once begins
to give off ramifying side branches at varying intervals
each having the same breadth as the mother channel.
Each of these subsidiary branches eventually divides
into a fan of terminal channels opening into one or
other of the several valli. Two of the main canals are
of special importance — those which diverging from
Magnavacca in the form of an immense V penetrate
ultimately to the Mezzano Valle and are the only means
of communication between this vast sheet of water and
the sea. These two canals serve roughly to separate
the eastern half of the lagoon into three divisions, a
southern section of three principal valli, aggregating
6,548 hectares (16,181 acres), a northern set of valli with
a superficies of 8,539 hectares (19,989 acres) and a
median division embraced between the arms of the V
with an area of 4,302 hectares (9,631 acres).

The Mezzano owing to the enormous area it covers
is the most important fish-farm in the lagoon, being 75
kilometres (46^ miles) in circumference, 14 kilometres (8f
miles) from north to south and 18 (about 10^ miles) from
east to west. Unfortunately it is the most remote of all
the valli from the sea, the shortest of the canals connect-
ing it with the Port of Magnavacca being 10 kilometres
long, while the southern sea-water sluices are as much
as 14 kilometres distant from the sea.

Command of large supplies of river water is also as
vital a necessity as communication with the sea both
in order to reduce excessive salinity within the valli and
as an integral factor in the annual stocking of the valli
with new shoals of fry. To effect this, six sluices fur-
nish communication between the river Volano and those
valli which approach the right bank of this stream; in the
same way the lagoon communicates at intervals along its
southern border with the Reno by five similar sluices.

The town of Comacchio lies at the present day in the
heart of the seaward half of the lagoon, upon a long

28

island connected with the mainland by a roadway carried
along the summit of one of the principal dykes.
Grouped round the town are the oldest of the valli, for
the seaward section of the lagoon, possessing ready
access to the sea, was naturally the scene of the initiation
and gradual evolution of the complex system now
existing.

To-day the town is 5 kilometres from the seacoast,
but I am of opinion from the character of the northern
portion of the lido or sea-bar that when Comacchio was
founded it was not more than 3 kilometres from the sea.*
At that time the valli now existing in the extreme north-
east of the lagoon, those between and including Valle
Volano and Valle Stefano, probably did not exist, the
sea-bar certainly was considerably narrower and access
between the lasfoon within and the sea without was much
more free that it now is — probably with a well-marked
and free tidal flow renewing the water of the lagoon
daily, making it a laginia viva or living lagoon such as
is paralleled by the condition of that of Venice to-day.
Now, the lagoon being dyked all round and with water
not in ample and free tidal communication with the sea,
it has become the type o{2,lagunamorte or dead lagoon,
in the lansruao-e of Italian valli-culturists— an inland sea
shut off from the Adriatic, save for the narrow sluices
opening from each valle into the marine canals.

The information we possess concerning Comacchio
does not antedate the 6th century, but there can be
little doubt that the principles governing the present
system of fish-farming were conceived and the foundations
of valli-culture laid several centuries earlier. The system
followed is in substance similar to that known to and
practised by noble Romans shortly before the beginning
of the Christian era as detailed by Pliny and other early
writers ; it is very probable that its introduction into the

* This conclusion receives confirmation from the fact stated by Feletti *
that the seaward extension of the sea-coast south of the mouth of the Volano
has been 20 metres in ten years. If we assign the founding of Comacchio
to the 5th century — say 1,500 years ago, this would give an extension of 3,000
metres as against the 2,000 which I claim. But of course the rate of deposit
is not and cannot be uniform and it is probable that the rate of deposit is
rather more rapid now than some centuries ago, as the extension of cultivation
and of drainage tends to make floods more violent and so bring down more
sediment and so increase the rate of the formation of deltaic lands. (*" Intorno
all attuale salsedine della acque del Campo Trebba, " etc., p. 3, Comacchio, 1909.)

29

Comacchio lagoon was due to Roman influence and
owed its inception there to the enterprise of men familiar
with the fish-pond culture of Rome and Naples.

We know little or nothing of its history during the
middle ages ; whoever ruled in Ferrara from time to
time doubtless dominated the lagoon, this being the
greatest of the feudal strongholds of the immediate
neighbourhood. In the 16th century the fishery in the
lagoon was one of the chief sources of revenue to the
Este family, Dukes of Ferrara. In 1598 Pope Clement
VIII dispossessed the Estes and seized both the Duchy
and the Comacchio lagoon. The claim for compensation
made by the Este family against the Papal Court stated
their annual receipts from the fishery at 300,000 lire (Rs.
1,80,000) but this was probably an exaggeration. At
this period, the end of the 16th century, the lagoon pre-
sented a very different appearance to what it does now.
In spite of much extensive reclamation made of recent
years the area at the period above named was consider-
ably smaller. Then the greater portion of the Mezzano,
now by far the largest of the lagoon fish-farms, did not
exist as such. It was a marsh choked with reeds and
rushes into which flowed the drainage of a great area
of cultivated land lying to the westward ; here this water
stagnated, percolating slowly through the separating
dykes into the adjoining valli. From its position as a
vast marsh between the lagoon proper and the cultivated
land it indeed derived its name Mezzano or middle.

In this condition the lao-oon remained till Cardinal
Palotta was appointed by the Papal Court as Governor
of Ferrara in 1631. Finding the sea-channel from
Magnavacca to Comacchio badly silted and capable
neither of admitting a sufficiency of tidal water for the
health and prosperity of the different valli nor of per-
mitting easy navigation to Comacchio, he had the channel
not only cleared and widened but extended beyond
Comacchio sufficiently far to give access to the northern
region of the Mezzano, a distance of 6 kilometres from
the city. This canal, known to this day in its seaward
section as the Canal Palotta, measures 6\ to 7J yards
wide with a depth of some 6 feet. By these means the
whole of the Mezzano was brought under control and
useless marshes were converted to fish-rearing purposes.

3o

Cardinal Palotta is ranked as one of the greatest of
Comacchio's benefactors and well may that be for under
his enlightened guidance the fish-rearing establishment
attained full development. On the completion of his
schemes the lagoon had practically the same area and
arrangement seen by Coste fifty years ago. Whether the
conversion of the Mezzano marshes into a vast and very
shallow fish-pond has been justified by experience is
however doubtful, for since Palotta's time there have
been repeated periods of great depression in the industry
caused by excessive and wholesale mortality among the
fish of the great ponds furthest removed from the sea;
the Valle Mezzano in particular being the one most
frequently the scene of frightful fish mortality.

With the exception of the sixteen years between 1 708
and 1 724 when the Austrian Government held possession,
the lagoon remained under Papal Government from the
end of the 16th century until the close of the iSth century
when Buonaparte in 1797 seized it during his campaign
against Venice. Bein£ loot he could not remove to
Paris, he wisely sold it to the Commune of Comacchio.
The ensuing sixteen years are perhaps the most notable
in the history of the establishment; they far exceeded in
continuous prosperity any other cycle we know of for
Pepoli * records that the average yearly revenue for
these years amounted to no less a sum than 567,749 lire
(Rs. 3,40,650) ; since then both the revenue and the
annual catch of fish have suffered cycles of violent fluc-
tuations. On the whole the condition and revenue of the
lagoon were not satisfactory during the last seventy-five
years of the 19th century. At present the lagoon as a
fishery establishment, is enjoying a period of renewed
prosperity due partly to climatic causes, partly to more
careful and scientific management, and partly to the
reclamation and utilization as agricultural land of large
areas of the least productive fish-farms.

Principles and Details of the System.

The principles governing the system of fry capture
and fish-rearing practised at Comacchio are essentially
the same as those forming the basis of the Arcachon

Fide Bullo, Piscicultura marina, part i, p. 365, Padua 1S91.

3*

system ; in both the aim is to utilize the natural instincts
of certain species of fish in such manner as both to stock
enclosed ponds automatically with an ample supply of
fry at one season of the year and at another season to
trap in quantity such fishes as have attained maturity
and whom instinct leads to attempt to pass to the sea for
spawning purposes.

The differences and modifications which we shall now
describe are clue in the main to the larger field of opera-
tions which is involved ; the one deals with ponds of
restricted area, the other with sections of an inland sea.

Apart from the matter of relative size, one of the
most important divergences seen at Comacchio is the
provision of separate openings in the encircling embank-
ments of each pond or valle for the entry of the fry during
ascent (monte'e or montatd) and for the construction of
pounds or labyrinth traps for the capture of the adult
fish. This separation of function is rendered necessary
owing to the great extent of the principal ponds and in
order to gain full advantage from the freshening influence
of the non-saline water of the rivers flowing along two
sides of the lagoon.

All the chief ponds communicate on the one hand
with these two rivers by means of massive sluices, and
on the other with the sea by one or other of the branch
canals arising from the sea-channels which traverse the
lido at Maenavacca or Bocca del Bianco. It is the aim

o ...

of this dual communication between the ponds and the
sea on the one side and the ponds and the rivers on the
other, to render possible the mechanical stocking and
fishing of the various fish-farms and at the same time to
introduce fresh supplies both of river and of sea- water to
moderate the high salinity which otherwise would be
caused by evaporation and generally to assist in main-
taining healthy conditions within the lagoon.

(a) The stocking of the Lagoon.

To give entry to the fry of those fishes whose instinct
leads them to enter river estuaries when very young, on
2nd February in each year, the sluices leading to the
Reno and the Volano are opened, together with all those
in the sea-canals leading to Magnavacca or the Bocca
del Bianco. A temporary third channel, the Taglio del

32

Gobbino, is opened at the same time through the sand-
banks of the Lido to provide direct communication
between the sea and certain of the southern valli and
carry thereto adequate supplies of sea-born fry. To
further facilitate the entry of fry and to ensure a uniform
distribution throughout the more remote valli as well as
those fed directly from the river sluices and sea-canals,
numerous gaps are cut in the dykes separating the valli
from one another and in the banks of the sea-canals
themselves. Every effort is directed to putting every
valle in free and untrammelled communication with the
rivers and the sea; to this end the sluices are opened,
the canal banks cut and dividing embankments breached;
all other duties yield to the vital necessity to obtain a
large stock of fry.

During the period the lagoon is thrown open to the
free influence of tidal flow, a distinct current is produced
during the ebb, flowing outwards through the river
sluices and through the sea-channels. As we have seen
when discussingthe Arcachon system, this outward current
is the important factor in attracting the fry of eels, mullet
and atherine smelts — the three species particularly
desired at Comacchio — towards and into the lagoon as
at this stage in their life the fry obey an instinct which
leads them to swim against any sea-going current ; when
they once enter the lagoon they scatter and show no
immediate desire to undertake further journeyings.

The abundance of fry entering the ponds being a
matter of the gravest anxiety to the authorities during
the months of ascent, employes of the commune supervise
the entrance of the shoals through the various sluices
and gaps. From time to time these men assure them-
selves of the quality of the harvest. The method is a
rough one, but answers sufficiently well ; it consists of
lowering bundles of brushwood into the entrance
channels ; these are periodically taken up and well
shaken or beaten, an action which causes all the eel fry
that have taken refuge in the interstices to fall out ;
according to the quantity thus obtained they gauge the
abundance or scarcity of the run of fry.

The fish whose fry penetrate into the lagoon in this
manner consist in the main of eels {A. vulgaris), smelts
(Atherina Boyeri), gobies, mullet of several species ; a

33

fair quantity of bass (Lad rax /uftus), soles and plaice also
enter together with various crustaceans and molluscs.
The young eels at the time when they enter estuaries
are 6 to 7 centimetres in length and it is estimated that
some 3,600 individuals go to a kilogram. The mullet
fry are still smaller, 20,000 going to the kilo. The
former attain maturity only after a residence of 4 to 5
years in the ponds when they should weigh from 4 to
5 lbs. each. During the period of ascent, February,
March, and April, care is taken to prevent net-fishing off-
the coast in the neighbourhood of the lao-oon, but after
the closing of the lagoon at the end of April or the
beginning of May, a large quantity of young mullet are
caucrht bv means of fine-meshed nets in the shallows
along the coast ; this particular stock is kept separate
from the general herd, being placed in special ponds.

When the inward migration of the young fishes
diminishes to inconsiderable proportions, usually about
the close of April, measures are taken to close all means
of egress from the lagoon. The river sluices are shut,
the gaps in the embankments are repaired and the
mouths of the permanent passages leading from each
valle to the adjoining sea-canal are closed by the con-
struction of palisades of stakes, to which are bound
tightly packed bundles of reeds, planted upright in the
mud. These particular obstructions are termed tresse ;
they occupy a most important position in the working
of a valle, for while they must be capable of impeding
the escape of fish to the sea, they must allow tidal
water to enter the otherwise closed ponds — a difficult
proposition.

The openings closed by tressa-palisadings are re-
stricted to those centres known as fishing stations where
permanent passages are maintained either between each
valle and one of the sea-canals or between adjacent valli.
In general each valle has one fishery station, but some
have more. Mezzano has 4, Campo and Ussanola 3,
Cona has 4, Vacca has 2. From each of these stations,
branch out towards the particular valle served, a number
of short canals called Covole (text-figures 4 and 7) which
have the function of reducing the tidal Mow received
from the main and wider sea-canal by dividing it into a
number of smaller streams and thus discharging it into

34

a valle by several mouths instead of by a single large
one. It is at the inner or valle mouth of a covola that
the tressa palisading is erected. Usually it takes the
form of a large rectangular enclosure projecting well out
into the water of the valle. When constructed at the end
of the stocking season, the palisading is complete on all
sides except that facing the covola. A simpler form
sometimes used is where the tressa has a V-shaped form,
being constructed of two palisades built out from the
mouth of a covola and inclined inwards or towards one
another so that they eventually meet in an acute angle.

OH T.mnii-rrTTTTTimill.imP

COVOLA

QUADRANGLE

OF

TRESSA

Pt'it.iiiii n li.l nil llllT

Fig. 5. — Plan of a true tressa.

PSEUOO TRESSA

VALLE

Fig. 6. — Flan of a pseudo tressa.
Earthwork shown by dotted shading.

Both these forms are designed to afford a very large
surface for percolation, just as the screens used in fresh

35

water fish-culture should be made several times larger
than the sectional area of the stream they intercept,
in order to allow for loss through the clogging of
some of the meshes and still be able to pass the full
quantity of water without difficulty.

The reed component in the tressa wall is the one
essential to the proper performance of this function ; the
other parts are designed to give necessary support. The
reeds are first tied into handy-sized bundles and then
built up around the mouth of the covola or passage into
a very tighty compacted palisade six inches in thick-
ness. The bundles are made sufficiently long to admit
of being firmly imbedded at the lower ends in the mud,
while the upper project 2 to 2\ feet above the normal
level of the lagoon. The method of supporting these
bundles by a frame work of upright stakes and hori-
zontal bars may readily be understood by examination of
plates V and VI, figures 7 to 10. As seen in these photo-
graphs these reed palisades vary in thickness and in the
strength of their supporting frames according to the
duties they have to perform ; in the majority the reed
hedge is staked on both sides (fig. 10, pi. VI), in others
a single lateral row of supports suffices.

When first built a tressa by its peculiar form offers a
very large percolation surface and the sea-fiood welling
through the feeder canals passes rapidly through the
reed hedge into the lagoon. Unfortunately the spores
of algae and the swimming larvae of various zoophytes
and other sedentary marine animals are likewise borne
up by the tide and settle upon the palisade ; here they
find conditions extremely favourable for their growth
and in a short time they fill the interstices so completely
that the walls of the tressa become all but impermeable ;
the volume of water passing through becomes so reduced
that it may be said to have no appreciable influence
upon the salinity of the lagoon thereby contributing
largely to the gravest danger to which Comacchio valli-
culture is exposed as we shall presently — (see p. 44).

When once the tresse have been built and the gaps
in the embankments repaired, work becomes restricted
to surveillance and the cutting and preparation of the
reeds and other materials that will be necessary when
the fishing season begins in autumn.
3-a

^

(b) Method of Fishing the Lagoon-ponds.

The capture of the fishes reared in the lagoons is on
a scale of such magnitude that ordinary fishing methods
do not suffice ; 1,000 tons of fish have often been taken
in a single season. The fishing period is usually con-
fined to the months of October, November, and Dec-
ember, when the instinct of the confined fishes, of the
adult eels in particular, is insistent for migration to the
sea. Last year the season opened exceptionally early
having already been some weeks in full swing at the
time of my visit on 29th September. The catadromic
instinct seems indeed to be developed considerably
earlier in the year in eels living in the Comacchio lagoon
than in those of the adjoining Venetian lagoons, where
the migratory instinct is developed in eels at the end of
autumn as in the case of these fishes in French and
English streams. Samaritani states that long experience
and many observations have demonstrated that in the
Comacchio lagoon, where the water is always very
different in density from that of the sea, especially during
the heats of summer, shoals of adult eels crowd around
the tressa palisades sometimes as early as the months of
June and July attracted thereto by the percolation of
sea-water through the hed^e of reeds. Indeed with
such energy do they attempt the passage of the barrier
that nets have to be placed in the channel behind the
tressa and it appears probable that many do actually
escape. As to the mullet, their instinct to make for the
sea begins to be active early in August.*

Usually the commencement of the great fishing
season is postponed till the end of September when
high spring tides and gales from the eastward may be
expected — two of the principal factors in inducing a
heavy run of fish into the fish-pounds. Several weeks
prior to what is termed the taglio delle valli, the
removal of a short section in each tressa palisade to
form an aperture of restricted width to permit the escape
seawards of the adult fish within the ponds, every
fishing station throughout the lagoon becomes the scene
of great activity. The whole staff engages upon the

* G. Samaritani. Prowedimenti necessari nella Laguna di Comacchio,
p. 18. Venice, 1899.

37

preparation of the materials required for the construction
of the great fish-pounds named labyrinths by Coste, in
which the fish are to be collected upon their escape from
the valle. This work having been completed at the time
of my visit to the lagoon, I shall fall back in the next
paragraph upon Coste for a description rendered in his
usual vivid manner, translating freely : —

"Every valle is transformed at this time into a combined work-
shop of wicker working and carpentry. All the employes are there
occupied, some in binding together bundles of reeds (Arundo
phragmites) into sections destined to form the walls of the labyrinths
whereinto the play of tidal water will attract the fish, the others in
preparing the posts which will support these permeable partitions.

"The bundles of reeds are made up into short hurdle or screen-
shaped sections about 4 feet long by some 7 feet high ; by overlapping
these sections and by binding them tightly together they are enabled
to form lengths of whatever extent may be desired while by super-
posing them like the leaves of book they are able to make them of
thickness suizable for the special service they are required to fulfil.
Each section is strengthened on one surface with two transverse
parallel rows of poles (or battens) which extend from one extremity to
the other, and vertically with two strong posts which project beyond
the lower margin in such manner as to be capable of being driven
into the earth.

" When they have completed a sufficient number of these sections,
they begin to place them in position, that is to say, they commence
to organize the labyrinths whose construction although very simple
exacts however so much precision and care that it is entrusted only to
the most experienced of the vallanti and sometimes even to an engi-
neer.* They adjust several of the reed sections end to end and thus
form palisades of length proportioned to the particular requirement.

Then making two great lengths of this palisading, they plant them
vertically in one of the passages at the extremity of one of the tidal
channels leading from the Palotta Canal. The anterior extremity of
one of the palisades is embedded in the left bank, the corresponding
extremity of the other against the right bank. These two walls being
thus secured in front — that is to say towards the end of the channel
where it enters the lagoon — the posterior extremities — those directed
towards the sea — are brought together in the centre of the channel in
such a way as to press against one another but lightly. In this way
two palisades thus disposed form a V-shaped angle whereof the
opening looks towards the lagoon while the vertex is directed towards
the (tidal) current which comes from the Adriatic.

" Matters being in this state and without changing the disposition
which I have just described, the work people drive into the soil the
posts with which the palisades are armed until the lower edges of
these partitions are pressed sufficiently deeply into the mud that

* For some years past the Administration have employed a highly skilled
engineer upon their permanent staff.

38

nothing can pass beneath. But as the mouths of the canals where
they construct these works are constructed in such a fashion as
to have only 3 to 4 feet of depth, it follows that the bundles thus
implanted and which are 6 to 7 feet high, project above the surface
(of the water) two feet at least and sometimes even 3 feet.

" If a fish which has left the lagoon passes now into one of the
mouths of the Palotta canal in order to escape to the Adriatic, it will
be compulsorily conducted to the extremity of the acute angle where
the two palisades touch without being attached to one another. There
if it makes an effort to pass beyond, these two walls yielding slightly
to the pressure, will open and then spring back into position after the
fish has passed between, closing the issue in such manner that it
becomes impossible for the fish to return through it. But it would
be free to gain the sea did it not find itself within a second com-
partment of the labyrinth where it thenceforward remains a permanent
prisoner. On passing through the narrow opening at the angle of the
first barrier, it finds itself in a heart-shaped chamber (botteghino)*
attached at its base to the outer surface of the two first palisades in
such a way that the acute angle of the latter barriers projects into its
interior. This small chamber, whose walls, formed also of fascines of
reeds, are supported exteriorly with stakes in similar manner to the
walls of the whole labyrinth, has also a narrow slit-opening at its apex
which permits the prisoners to advance into the more distant
chambers. Beyond the small botteghino are found two other palisades
disposed in an angle of which the base, in the manner of those first
described, includes the whole width of the canal and whereof the
apex, directed against the tidal current, gives ingress by means of a
narrow opening into a second small chamber into which the fish
eventually pass after traversing the space bounded by the second
series of palisades. Arrived in this second confined space (baldresca)
not only are the fish unable to retrace their way but not finding any
opening at the further extremity, as the walls are continuous they
remain there captive if a mullet, a sole, a bass or a sparoid, for these
species are not lithe enough to traverse the small openings in the
barrier ; if on the contrary it be an eel, it insinuates its head or its tail
between the reeds and with the help of the vigorous movements of
which it is capable it slips through the walls of its prison leaving
behind it all those of its companions whose form is not adapted to
such muscular exercise. But this exhibition of skill and strength
does not help it to freedom. It falls into yet another triangular
space, vestibule of its definitive prison, where after wandering a
longer or shorter time without succeeding in traversing the wall of
which the thickness and strength have been specially calculated to
resist its endeavours, it finds only three exits similar to those through
which it has already passed and which are situated at the angles of the
triangular space where all its efforts to escape have been checkmated.
Finding no other passages practicable and tired of the struggle, it

* The shape of this first small chamber [botteghino) in those fishing stations
which I inspected is actually diamond-shaped. The only heart-shaped chamber in
the labyrinths as now constructed is the baldresca where the mullet are separated
from the eels.

Fig. 7.— Plan of a Lavoriero or labyrinth, showing its arrangement and its relation

to the other tarts of a fishing station.

A. Botteghino. B. Kaldresca. C. Coloura vivo.

D. Cogolara. O. Otelior eel-trap chambers.

E. Opening in Tressa palisade to give free communication with the valle served by this

labyrinth. Land and embankments shown by dotted shading.

39

finishes by passing through one of the openings available. But.
behind each of these issues, a final chamber with walls as impassable
as those of the large triangular space it has quitted, yields it into the
hands of the fishermen.

" These ingenious devices which the Adriatic currents permit of
execution, are not therefore limited to the attraction of fish from the
lagoon into the sea-channels ; they assist as well in the separation
of the species just as the mechanism of certain factories separates the
material which is the object of their operations. The art of fishing is
elevated here to the height of an industry which reposes on principles
of which the application conducts to a termination contemplated in
advance and always with identical results. This industry marks the
place where the crop shall be gathered and each species proceeds to
the store compartment which is assigned to it ; there is only to open
a sluice in order to operate this marvel which during the space of
three months collects yearly the matured harvest of the lagoon."

The construction of a fishing labyrinth, termed lavor-
iero at Comacchio, will be readily understood from the
foregoing description read in conjunction with the
diagram here inserted (fig. 7).

A fishing labyrinth divested of detail may be described
in short as an arrangement of vertical walls made of two
or of three thicknesses of reed screens embedded in the
bottom of a sea-channel and supported by vertical posts
and horizontal poles, the whole arrangement forming a
succession of angular enclosures, each having the open
base turned towards the passage {covo/a) leading to the
lagoon, and the apex towards the sea. The fish enter
the main or second enclosure, the lavoriero proper,
through a narrow vertical opening cunningly contrived
at the apex of the first series of palisades. When in turn
the fish attempt to leave this principal chamber, in passing
out the species separate automatically in such manner
that the mullet are imprisoned in a small round or heart-
shaped pound, called baldresca, at the apex of the palisades
of the main enclosure, while the eels pass through the
comparatively thin wall of the baldresca into a terminal
triangular enclosure havino- a small chamber or otele at
each of its angles (pi. V., fig 8, and pi. VI., figs. 9 and 10).
The walls of the otelia.s well as those of the large enclosure
whereof they form part, are constructed in a most massive
manner ; the reed screen, composed of a great number
of layers of reed fascines, is usually from 7 to 8 inches
thick supported by stout piles and secured by strong
horizontal baulks of timber on the exterior as seen in
pi. V., fig. 7, and pi. VI., fig. 10. From the final traps

40

the fish are lifted by means of a dip-net, and placed in
great spherical wicker store-pots till removal to the
preserving factory.

This so-called fishing of the lagoon continues usually
during a period of three months. The amount of fish
collected each day in the various pounds varies with the
state of the weather. Large catches occur only on
moonless or overcast nights ; when storms from seaward,
especially the levante-sirocco blowing from E.S.E.,
coincide with the occurrence of such black nights, the
quantity of fish that accumulates in the lavoriero is often
so exceedingly great that the inner pounds become
choked with a solid mass of struggling fish. Bonaveri*
records that after one such night (4th October 1697),
322,520 kilogrammes of fish were taken from the pounds,
as much as 64,504 kilos being yielded by one valle alone.
Ouantities so vast are inconvenient to handle, so when-
ever the workers find the fish accumulating too quickly
in the enclosures, they light bonfires at the sides of the
tressa and this has the effect of checking the migration
temporarily.

The reason for the great rush of fish into the lavorieri
during a tempest is that when this is from the eastward
it raises the water unusually high in the tidal-canals and
thus induces an increased flow of water through the
palisade barriers into the lagoon — a call which the adult
eel and mullet hear and act upon immediately. Occasion-
ally at these times, the storm heaps up so great a
quantity of water in the canals, as to entail such pressure
upon the walls of the labyrinths as to compromise their
stability. To avoid this, although the proceeding tells
against the success of the day's fishing, it becomes
necessary to stop the flow of water through the palisades
into the lagoon by shutting certain sluice gates which
are provided for this purpose either in the covole or in
the sea-canals themselves.

It is obvious that during the summer season when
the walls of the tresse are continuous, and also during the
fishing season when, although the tresse are open, the
terpiinal (inner) mouths of the sea-canals are completely
obstructed by the palisades of the labyrinths, special
means must be provided to permit the employes and

* Storia della Citta di Comacchio, 1761.

4i

watchers to move freely in their boats from valle to valle.
Accordingly we find at almost every fishing station, a
narrow service-canal passing round at one side giving
access to the valle from the feeder canal from the sea. In
other places a similar narrow boat canal cuts the separat-
ing embankment between adjoining valli. To prevent the
escape offish by way of these service-canals, wooden locks
of the simplest description are installed in convenient
positions in their course. Figure 12, pi. VII., depicts
clearly the appearance and arrangement of one of these
locks. The lock " gates " are stoutly built wooden
shutters extending the width of the cut, the ends and
lower edge fitting into wooden grooves in the wood
frame built into the bank and bottom of the channel.
The lower edge of the lock-shutter is shorter than the
upper to facilitate removal and replacement in the
retaining grooves.

^.j!ii^LiL-.'^~ »

Fig. 8. — Diagram of wooden shutter used as a lockgate on the
service-canal between a sea-canal and a valle.

A, Handle by which it is pulled up when the lock is to be opened.

Cycles of Bad Seasons ; Causes and Suggested

Remedies.

From time to time in spite of the great store of
empirical knowledge gained from centuries of struggle,
there occur disasters in this industry which impair its
prosperity for a series of years. During the past century
no less than six years are recorded as marked by exces-
sive Joss or mortality of fish, entailing a succession of
unremunerative fishing seasons varying in degree accord-
ing to the severity of the initial cause. It is probable,
I believe, for the reasons to be stated later, that these
cycles of disaster have tended to occur at shorter
intervals and with great intensity since the Mezzano was

42

added to the fishery system in the middle of the 17th
century ; at no time however can this lagoon system
have been immune since the period when it ceased to be
a lag una viva — one wherein the sea daily ebbed and
flowed freely throughout its whole extent. Blizzards
from the Alps, exceptional floods in the rivers bounding
the lagoon, intensely hot and dry summers — these are
the three factors which are recognized as the principal
causes of seasons of failure. For example, in 171 1
a tempest of such intensely cold wind descended upon
the lagoon that the fish died in myriads. This followed
upon other losses in 1705 and 1709 caused by flood
irruption of the adjoining rivers, whereby the dykes and
tresse were submerged, allowing a great proportion of
the imprisoned fish, young as well as mature, to escape
to the sea.

Again, in the years 1 7 1 8 and 1 7 1 9 the heats of summer
were so intense and long-continued that a great portion
of the lagoon dried up and immense deposits of salt
formed upon the bottom. Bonaveri states {loc. cit.) that
salt to a value of over 18,000 scudi was collected.
Owing to the immense amount of evaporation in these
years the water that remained in the deeper sections
became so highly saline that widespread mortality
resulted amono- the fishes.

These disasters were exactly paralleled in the course
of last century. In 1825, 1834, 1890, 1 891, the summers
were abnormally warm and as a result of active evapora-
tion over the 40,000 hectares of shallow ponds the
resultant high salinity of the water entailed enormous
loss among all ages and species of the fish in most of
of the ponds in the lagoon. In 1890 the seaward flight
of the fish into the labyrinths was a veritable stampede
and disastrous in proportion. The catch surpassed a
million kilogrammes (say, 1,000 tons) in great part com-
posed of minute and immature fish such as required a
sojourn of several years longer in the ponds to attain a
proper size. In the pounds of a single valle, that of
Vacca, 100,000 kilos of fish were taken, while Valle
Fossadiporto gave half this quantity within a few days.
Much of the fish taken in the labyrinths was so tiny
that the greater part after being placed in the great
wicker store-baskets passed easily through the interstices
and so escaped to the sea.

43

In the succeeding year there was a similar scarcity
of rain, and the inimical conditions being' intensified the
slaughter of 1890 was repeated in those ponds which
had been fortunate in escaping the dangers of that year.
As a d'rect result of the excessive mortality characteris-
ing 1 890- 1 89 1, the catch of fish during the ensuing
seven years was abnormally poor and the Commune lost
over a million lire (Rs. 600,000) on the working of the
lagoon during this period — a large sum for a town whose
prosperity is dependent almost entirely upon the produce
of this fishery.

The enormous loss entailed by the winter storm of
171 1, was repeated in 1850, entailing a series of
miserable harvests during the next cycle of 7 years,
while the floods of 1705 and 1709 had their counterparts
in 1859 and 1862, when the Reno burst its banks and
flooded the lagoon, liberating multitudes of immature fish
and causing the fisheries of 1863- 1865 to be among the
poorest on record.

The followino- table onven bv Samaritani, * to whom I
am indebted for the majority of the figures above quoted,
shows clearlv the fluctuations in the average catch during
alternating cycles of prosperity and poverty :—

a

V

>»

0 6

Catch of eels and mullet.

In the first

In the last

Total for

Annual

Remarks.

' 0

0

s °

year of
each

year of
each

each

average for
each

0)

3 0

period.

period.

period.

period.

kilos.

kilos.

kilos.

kilos.

1798-1824

27

894,168

i,245>l63

29,579,327

1,095,530

1825-1838

14

487,977

585,207

6,293,110

449,508

Great mortality
due to high
salinity.

I 839- I 849

1 1

761,632

682,447

7,222,767

656,615

A winter storm

IS50-I857

8

571,290

484,610

3,567,023

445,878

in 1850
caused much
mortality.

1 8 58- 1 860

3

613,548

715,509

1,974,783

658,261

Lagoon flooded

1S61-1S6S

8

455,9o8

543,265

4,183,121

522,890

from Reno in
1859 and
i860.

1 869- 1 890

22

639,060

1,024,257

19,915,987

905,272

High salinity

1891-1898

8

365,028

444,233

3,498,<'77

437,26o

in 1890 — 1891
caused great
losses.

Loc, cit.

p. 7.

44

Besides the three main causes which periodically
entail heavy losses on the valliculturists of Comacchio,
there are several others which are either subsidiary and
intensifying in their influence or which may be limited
in their evil influence to certain sections of the lagoon.
Of those latter Dr. E. Feletti mentions* an instance of
one valle (Trebba) where percolation through the
embankment dividing it from reclaimed land lying at a
lower horizon, dangerously reduced the level of the water
within, so producing a strong indraft of water from the
maritime canal, which had the effect of causing a stam-
pede of fishes into the labyrinths. Another cause of loss
he traces to the liberation of poisonous gas (sulphuretted
hydrogen) from the bed of the various chambers of the
labyrinths, when from any cause the fish crowd in in
excessive multitudes. Probably this is more or less true
and the remedy is obvious that these passages should be
cleaned out periodically and mud foul with accumulated
organic matter removed. I would point out however
that the mere disturbance of a mud bottom, apart
altogether from any liberation of asphyxiating gases, is
enough to kill many species of fish, mullet in particular.
And whenever the labyrinths are choked with fish jostling
one another and burrowing excitedly in the mud of the
passages, this danger must arise. The factor of poison-
ous gas will certainly also be present under such
circumstances, and this in combination with mud in
suspension may cause a great proportion of the fish to
die before it is possible to remove them from the
pounds.

Floods may be guarded against by raising and
strengthening the protecting embankments, and loss by
percolation implies faulty workmanship. Frost and
blizzards aQ-ain are to be fouo-ht and their ill effects
minimized by recourse to the device of excavating deep
pits and channels in suitable places in the ponds,
whereinto the fish may retreat when necessary. Bullo f
mentions that these refuge-canals {canale raccoglitore)
were first employed in 1863 and that experience has

* " Intonio all attuale salsedine della acque del Campo Trebba^ ed alia motia
tli anguille nelle sue Covole " ; Comacchio, 1909.
f Lot. <//., p. 406.

45

proved them to possess the greatest efficacy in protecting
the fish from the harmful influence of winter storms.
These refuges are always dug upon the windward side
of a valle.

Apart from the intermittent factors which from time
to time work havoc on a grand scale there appear to be
a number of causes at work tending towards a gradual
deterioration of the lao-oon as a fish-rearino- establish-
ment. Fifteen centuries and more have elapsed since
man began to build dykes to limit the free access of the
sea to this great river backwater and to hold up water in
the marshes of this delta. The passage of a period of
this length, short though it be in geological history, is
long enough to produce effects of importance under the
conditions found at Comacchio. Deltaic accumulations
may be slow in certain cases, but the change of estuarine
swamps into low alluvial land is none the less inevitable.
Thus we find that the depth of water in the inner
sections of the lagoon, those margining the western shore,
is gradually becoming less, as silt brought down during
rains by the channels that drain the adjoining land,
gradually accumulates upon the bottom. As a conse-
quence the water capacity of these sections becomes
reduced and when an unusually hot and dry summer
supervenes, the consequent loss of water due to evapora-
tion entails much more serious results than in former
times — with a reduced volume of water at the beginning
of the season, evaporation produces a salinity much
higher than it would, were the orio-inal volume o-realer.
So far has this silting of portions of the landward
sections proceeded, that extensive reclamations have
been made. Valle Gallare and Valle Voile, two of the
largest valli after Mezzano, have been wholly reclaimed
since Coste's visit in the middle of last century and
extensive sections of the Mezzano and of Fossadiporto
have also been turned into agricultural land within recent
years.

Supplies of fresh water to correct the tendency of the
enclosed waters to acquire too great a degree of salinity,
are in the main derived from the two rivers borderino- the
lagoon, the Reno and the Volano. These unfortunately
contribute less water than in former days, partly because
increased cultivation in their basins takes up greater

46

quantity and so reduces the flow, and partly because the
reclamations effected have entailed the loss of certain of
the old sluices. This latter cause has operated in a
specially adverse manner upon Valle Mezzano, the
greatest section in the lagoon, for to-day it possesses but
a single sluice admitting water from the river Reno
instead of the three it once had.

While the fact cannot be disputed that the evils
which result from certain defects in the present physical
condition of the lagoon may be mitigated in large part
if freer commumcation be established between the laeoon
and the sea and if larger supplies of river water can be
made available, the methods of so doing have been the
subject of much controversy. The present is not the
place to discuss the details of the arguments, as their
interest and bearing are almost entirely local and special
to the conditions of this particular lagoon. Additional
river sluices and better regulation of those existing are
among the most obvious remedies, but the river supplies
are limited because of several considerations, some of
which have been alluded to above, and permanent relief
can only be obtained if the supply of sea-water be enor-
mously increased. At present the lagoon is virtually
cut off from free access to the sea during the period from
May to the end of January. From May to September,
the massive palisades of the tresse form all but imper-
meable obstacles, while during the fishing season between
September and January the equally massive walls of the
labyrinths offer as effective barriers to the tidal flow.
Besides, what is gained by the lagoon during the flow is
lost in large measure during the ebb. Either some
radical alteration must be made in the methods of
penning in and of trapping the fish, whereby the various
barriers may permit free flow and ebb at least during the
dry season, when evaporation is intense — a proposition
no one has solved yet, or else means must be adopted
to greatly increase the quantity of water brought in
by the flood and then to prevent its reflux. The
difficulty of passing sufficient water in during flood-tide
by means of the restricted channels which alone are
available is, however, practically impossible so long as
the present method of constructing labyrinths and tresse
is followed.

47

Bullo has suggested the destruction of the dykes
separating the smaller seaward valli in order to form a
huge reservoir, whence sea-water might be passed into
the great landward valli. Samaritani, an engineer
whose practical knowledge of the conditions are unri-
valled, traverses the efficacy of this latter proposal point-
ing out that, as the evil to be averted is a rise in the
salinity of the lagoon water, admission of sea-water is
much less efficacious than that of fresh water.

Several of the best informed experts upon vallicul-
ture believe the conditions now existing in the lagoon
and in the lands borderino- it to be such that it is ad-
visable to reclaim and give to agricultural purposes the
whole or the greater part of the lagoon. Samaritani
gives figures * based upon the results of reclamations
already effected to demonstrate that this step would prove
much more profitable to the communal authorities than
to utilize the lagoon for fUn-farmingf.

Elsewhere t he suggests as a less radical measure
that, failing the adopt'on of a scheme for the total
reclamation of the lagoon, great benefit would result
to the remainder, if the less remunerative sections be
recla;med. If such were to be done, the area of the
lagoon might then be so reduced that the river sluices
now existing would be capable of supplying sweet water
in sufficient volume to keep the remaining ponds at
a suitable level and of a salinity agreeable to the fishes
inhabiting them.

The DisrosAL and Curing of the Products.

An average taken over a period of 86 years in last
century shows the annual catch of eels and mullet to
bej:—

Eels ... ... ... ... 695,818 kilos.

Mullet 78,43! »

Besides these a valuable amount of less important
fish, chiefly smelts and gobies, are caught and disposed of.

* Fide Bullo, loc. cit., p. 365 et seq.

f Provvedimenti necessari, etc., p. 96.

% Samaritani, quoted by Bullo, loc. cit., p. 366.

43

In a good year about 90,000 kilos of smelts may be taken
and 2,500 to 3,000 kilos of gobies. *

Fully two-thirds of the eels and smelts are cooked
and preserved in vinegar — marinated. The rest is sold
fresh, except for special orders and under exceptional
conditions of superabundance, when such large quantities
are caught as exceed the resources of the marinating
factory and of the receptacles for live storage.

Under present arrangements the whole produce of
the lagoon is leased out by the Valli administration for a
term of years, eight in the contract now running. Under
the terms of this agreement the Commune, as the owners
of the fishery, retain full control of the fish farms, employ-
ing their own permanent staff upon all the necessary
duties, including that of collecting the fish. The lessees
have nothing to do directly with the working and
administration of the Valli, the administration supplying
loads of fish as caught from day to day during the season
at the lessees' factory.

At the time of my visit the lease of the produce was
held by Messrs. Anselmo Cornia &. Co., and Signor
Cornia was kind enough personally to conduct me over
his works and to explain fully the processes in operation.
The factory is new and most conveniently designed. A
canal runs along; the rear with sidinQ-s branchinor into
the building itself. In this way, the loads of fish are
brought by boats without handling or delay right into the
the hands of the workpeople.

I should here premise that the method of curing is
limited almost entirely to the marination of eels and
smelts, a process said f to have been invented (more
probably improved) by Count de Bonneval, a French
soldier of fortune, who commanded the Austrian troops
who seized Comacchio in 1708 and who remained there
for some years in charge of the garrison. Bonneval
appears to have been a man of versatile talents and keen
to do his utmost to advance the prosperity of the loca-
lity under his charge.

At the time of my inspection of Messrs. Cornia's
factory, work was in full swing. Aflat had just come in

* Goein, La Pisciculture en eaux salees, p. 212, Paris 1891.
f Bellini, La Laguna di Comacchio otti il dominio della Corte di Vienna,
p. 10, Venice 1908.

49

filled with a writhing mass of thousands of eels all in
the characteristic livery assumed when they are fully
adult and ready for their catadromic migration. The
boat lay in a little basin at one end of the heading room
wherein were a number of " headers " seated before
wooden blocks and each surrounded by several baskets.
The eels in the boat were scooped up by means of land-
ing nets, filled into baskets, and distributed among the
headers whose operations consisted in chopping off the
head of the still living fish and dividing the body at
intervals so as to give lengths of about a.¥ to each.
Only at alternate intervals was the body cut right
through ; at the other points it was merely nicked deeply.
This allows the halves of each double section to be
readily folded upon themselves at the joint made by the
deep nick at mid-length. As the men work they sort
the pieces as they go, flinging" them as cut into one or
other of the baskets disposed within their reach, accord-
ing to the size and quality of each section. Thus the stout
sections from the thick portion of the larger fish go into
one basket, the thin tail sections into another and
a third will take those of fish of inferior size. The heads
are considered as offal and are sold by the heap to the
workers and local people at a trifling price.

As the baskets are filled with the severed sections,
they are carried to the roasting room, where they are
first spitted and then cooked before great fires laid in
open fire-places. This roasting room extends the full
width of the building ; the fires, eight in number, occupy
the whole of one side while along the side opposite are
the spitting tables. The floor-space between the fires
and the spitting tables is occupied with wooden racks on
which the filled spits rest before and after roasting.

Fig. 9. — Eel-roasting spit showing method of charging with
double lengths of eel-sections.

The spits upon which the double sections of fish are
transfixed are iron rods about 7 feet in length provided

50

with an L-shaped handle at one end, so that when the
rod is suspended upon a bracket at either side of the
fire-place it may be turned or rotated by hand. As shown
in text-fig. 9 the alternate pairs of eel sections are
roughly arranged with their lengths at right angles to
one another with a view to effect thorough cooking by a
free exposure of the maximum surface. Eight of these
spits are suspended on brackets before each fire, the
whole number being arranged in a single vertical row.
The four upper spits are tended by one woman, the four
lower by a second ; both stand at the right-hand side of
the fire-place, as the handles are all placed on that side.
These two workers attend solelv to the roastinq- of the
fish — to the proper rotation of the spits. On the left
of the fire other women see to the placing and removal
of the spits. The fresh spits with the raw fish are placed
on the upper brackets ; the lower rows get the greater
heat and as they are completed and removed those
above are moved down to occupy their places. The
spits with cooked fish when removed from the fires are
hung on the racks in the middle of the room to cool
awhile before other workers strip the fish from them.
Wood fires are used, laid on a flat open hearthstone.
They are well supplied with fuel and are kept blazing
furiously, giving out an intense heat. As the fish roast
before the fires, a considerable quantity of oil oozes out.
This falls into a sloping gutter in front of the hearth
and flows along it into an underground oil-tank or col-
lecting reservoir. Some further oil is also collected
from the spits as they cool on the racks each of which
is furnished with a wooden trough at its base to catch
the oil drippings. The oil thus collected during the
cooking of the eels is employed in the frying of smelts.
When the roasted eels have cooled sufficiently, they are
stripped from the spits and arranged neatly in circular
shallow baskets which are then carried outside and laid
in rows in the shade of a wide verandah where they are
allowed to cool thoroughly before further handling.
Considerable importance is attached to this item in
preparation.

When quite cool the fish are carried to the packing-
room where the sections are neatly packed in wooden
barrels and kegs. The lower layers are packed fairly

5i

tightly, the upper ones rather less so ; when full the
barrels are headed up, and spiced vinegar containing a
little salt is poured in through the bung. The barrels
are left to stand thus for some time, additional vinegar
being added if necessary to completely fill the barrel,
when the time comes to close the bung. With proper
packing 12 litres of vinegar are calculated to suffice to
marinate 13 kilos of eels while half a kilogram of salt is
the amount added to each 4 litres of vinegar.

The barrels and kegs employed are made on the
premises. On an average about 22,000 are required for
the year's pack. The great majority of the marinated
eels are put up in four sizes of barrels containing respect-
ively 15, 25, 35 and 55 kilos offish. Austria and South
Germany are by far the principal customers and Signor
Cornia informed me that the value of the export of this
product from Comacchio to these countries may be
reckoned 200,000 lire and 150,000 respectively. Small
quantities also go to France and Russia. During the
middle ages these kegs strapped upon pack horses must
have been familiar objects to travellers on the historic
road between Venice and Augsburg, the great artery of
commerce in those days between Germany and the mis-
tress of the Adriatic. As showing the extensive nature
of this trade at the present day it is interesting to note
that the quantity of vinegar required during the three
months, which constitute the main fishing season, averages
350,000 litres, all made from grapes. Marinated smelts,
called in the Italian trade Aquadella fritufa marinata,
are treated somewhat differently from eels, partly because
of their smaller size and more delicate nature and partly
because of a deficiency of oil in their tissues. Instead,
therefore, of being spitted and roasted, they are first
rolled in flour and then fried or rather boiled in oil.
The oil used is that collected from the eels during roast-
ing ; in this they are boiled for five minutes in copper
caldrons, the flour in which they were rolled imparting an
appetsing appearance and colour to the finished product.
After frying they are laid out to cool and to drain
thoroughly. As with eels, after they are cool, they are
packed in barrels, which are filled with vinegar after
being headed up, and then left to stand awhile before the
bung is inserted.
4-a

52

A small quantity of the marinated fish is packed in
tins of various sizes, but this method does not appear to
be greatly developed as yet. As already mentioned
approximately two-thirds of the eels caught are mari-
nated. The remainder is sold alive, special live cars,
often of large dimensions being employed to convey
them often to distant cities along the coast and up the
River Po. They have the form and appearance of a flat-
bottomed lake boat ; they are sharp at either end and are
fully decked over. The sides and bottom are perforated
with rows of long narrow slits thus — O^— ^O- They
have no internal partitions and have but a single square
hatchway closed with a sliding cover capable of being
locked and well secured. The live cars, when prepared
to convey quantities of live eels, are corded round with
many lashings of rope to enable them to stand the strain
of towage.

Similarly shaped craft {marotd) but of greater size
are employed in large numbers as storage receptacles
whenever very large catches of fish are made in the
labyrinths. If the marinating factory is to be run
economically, it is obviously essential that a steady supply
of fish shall be available during the working season,
hence the value of these great store chests. A usual
size is 1 5 metres long by 3 metres beam (49 x 9f feet),
with a depth of 2 feet ; each fishing station has a
number stored away in its store-sheds ready for use and
as the eels live for days and even weeks in them without
harm, the administration is prepared to take full advant-
age of those notable niohts when Qreat multitudes of
eels simultaneously develop an instinct which urges
them to seek a way to the sea. In a single night as
much as 20,000 quintals (about 200 tons) weight of fish
may be taken from the pounds ; the factory requires
some weeks to dispose of such a quantity, so were it not
for the employment of many marote, work would be
forced at a wasteful pace and much of the fish would
otherwise be lost.

Besides these wooden boat-shaped store recept-
acles, numbers of spheroidal wicker pots of large
dimensions, 6 feet to 7 \ feet diameter, are used for
the same purpose. When comparatively small quanti-
ties of fish have to be stored, these wicker pots are

53

preferred to the larger and more cumbersome wooden
live-chests.

Eels and mullet are also occasionally prepared for
market by smoking' and whenever phenomenally heavy
catches are made, as for example during an excep-
tionally dry season when the excessive salinity of the
lagoon drives all species and all ages of fish to attempt
to escape therefrom, then resort is had to the salting
and sundrying of such fish as are not suitable for mari-
nating and of those quantities which are beyond the
capacity of the storage receptacles. The preparation of
salt fish under these circumstances is an exceptional
expedient resorted to only to save the fish from being
utterly lost. Except under stress of necessity this
method of preparation is not practised as such fish
brings in poorer prices than does either live or mari-
nated fish. The last-named is the favourite as it is the
only method of curing known at Comacchio which
gives better returns than the sale of live fish, 1 10 lire
being a fair average price for ioo kilos of marinated eels
as against 80 lire for the same weight in the fresh
(living) state. Mullet are worth fresh about 60 lire per
100 kilos, smelts 30 lire, and gobies about 21 lire.

Administrat 1 ON.

Since this great fishery establishment became the
property of the Commune of Comacchio, it has been
administered for the general good by a communal or
municipal committee who arrange all matters relating to
the upkeep, policing and fishing of the various sections
of the lagoon.

The communal administration may also undertake
the preparation for market of the produce of the fishery,
but under ordinary circumstances the total produce for
a term of years is offered for public tender at periodical
intervals and it is only when no satisfactory contract is
possible that the commune organizes a manufacturing
department charged with the curing and wholesale distri-
bution of its products. Such an eventuality arises only
after or during one of the periodically recurring crises
which afflict the industry when a succession of abnormally
dry seasons entails wide-spread death and loss of fish ;
the length of these cycles of poor seasons being indefinite,

54

commercial firms naturally hesitate to take the heavy
risks involved by a tender based on normal returns and so
discount the risk by heavy reductions in the sums they
offer for the produce of the lagoon. Cycles of unremu-
nerative years usually fall to be worked by the commune,
whereas during" cycles of prosperity, competition for the
fish-contract is usually satisfactory.

Each large valle or group of smaller ones has its
separate staff of workmen and fishers presided over by
an experienced chief, \\\z. fattore . These men are accom-
modated with quarters in the fishery building found at
each of the principal fishing stations. Here they live
and mess together like soldiers in barracks, a necessity
consequent upon the fact that the chief work of the year,
the herding of the fish into the labyrinths, takes place
during the night. The life is hard and trying but custom
and the traditions of centuries of such conditions render
the men passively content, satisfied if they be allowed a
couple of days, Saturday and Sunday, on alternate weeks
to revisit their families in Comacchio. To permit of this
arrangement half their number take holiday in alternate
weeks.

The working staff struck me as a fine body of men ;
true handy men they are, for their duties include the
supervision of the entry of the fry from the sea during
the spring, the repair of the dykes, the weaving of the
reed screens, the construction of the labyrinths and
sluices and the collection of the fish. A ration of fish
is given them in addition to wages. Repairs to the
embankments and fishing stations are executed during
the hot season (June, July and August) when the water
in the lagoon is low and other duties are light.

The work of policing is done by a separate staff of
over one hundred men who patrol the lagoon in light
boats both day and night. Their duties are no sinecure
for there are many poachers in the villages around the
lagoon and even in Comacchio itself. Coste stated his
belief that fully as large a quantity offish is stolen by
these poachers, as is fished by the administration.
Seeing that the perimeter of the lagoon is 177 miles, the
opportunities for theft are undoubtedly very great, but
while admitting this, I cannot believe that such an
enormous loss of fish is possible ; I incline to think that

55

Coste in his desire to exhibit the Comacchio establish-
ment as a model of everything excellent in marine
fish-farming, took this means to explain away, in part, the
extremely poor yield of fish per acre of water surface as
exhibited by the official statistics.

The central administration has the technical advice
of a trained engineer who supervises the general
mechanical arrangements, particularly the management
of the sluices and the reclamation of land from the
lagoon. Of late years it has become evident that
permanent deterioration of the fishery can only be
prevented by the assistance of science, and so the
engineer to the administration is now required to have
biological knowledge as well as an engineering training.
The late Engineer, Signor Samaritani, appreciated the
absolute necessity for a thorough biological knowledge of
the laooon and its inhabitants and laboured long but
without success to gain the assistance of Government for
the establishment of a biological station in the lagoon.

In cases of sudden crisis, the administration follows
the excellent custom of calling to its council not only its
engineer but also the working superintendents (fat tori)
of the different valli, and any old servants whose length
of experience entitles them to consideration.

Revenue.

The annual net profit derived from the fishery has
been and still is subject to enormous fluctuations owing
to the alleged impossibility of warding oft the evil effects
of series of abnormally hot dry seasons which occur
irregularly from time to time. Fortunately the cycles of
prosperity have usually been much more prolonged than
those which are not remunerative, and there is every
reason to believe that the years of unremunerative fishing
will be largely curtailed now that the. empirical methods
of the past are being submitted to the touchstone of
science, which by experiment and deduction has already
effected a measure of amelioration. When Pope Clement
VIII in 1598 seized the lands pertaining to the ducal
house of Este, Comacchio was included, and this family
claimed that they derived annual revenue from the fishery
of 500,000 lire. These figures are possibly exaggerated
as in 1587, it appears that the Estes had farmed it out

56

for 333.000 lire.* Be that as it may, after Rome seized
the fishery, its value began to decline ; from 1725 to 1748
the average net revenue was only lire 74,480. Thence-
forward it gradually recovered, being lire 104,931 for the
years 1749 to 1757, while from 1767 to 1797 it rose to an
average of lire 322,640.! At the end of the last-
named period, the fishing rights in the lagoon were sold
by Napoleon I to the Commune of Comacchio. The
ensuing period of 16 years, if we can trust Pepoli's figures,
mark the high tide mark of prosperity in the history of
the lagoon, for the average net revenue from 1798 to
1813 inclusive is given as lire 567,749 per annum. The
figures for the next period, 18 14 — 1825, show a large
falling off, the annual net revenue varying between lire
200,000 and 250,000. 1825 marks the year of greatest
disaster in the history of the establishment ; almost the
whole stock offish throughout the whole lagoon perished
through the partial drying up and consequent extreme
salinity of the water. The Government thereupon was
forced to resume the fishery and held it till the end of
1868 when it was handed back to the administration of
the local Commune. During this period the Government
paid a sum varying between lire 89,000 and lire 75,000
per annum to the Commune as compensation for occupa-
tion of the fishery, but the bargain was a bad one for the
Government as five separate disasters, in 1825, 1834,
1850, 1859 and 1865, entailed an average annual loss of
246,408 lire. J Strangely enough, the return of the
fishery to the direct governance of the Commune marked
once more the beginning of a long period of prosperity,
for, during the 22 years from the end of 1868, the net
annual revenue derived from the leasing out of the
produce varied between lire 121,318 and lire 185,000.
The hot summers of 1890-91 entailed great loss and for
a number of years the Commune being unable to lease
the fishery, had to conduct it at great loss. With the
close of last century recovery was made and it was again
possible to lease the catch of fish at a remunerative rate.
The fishery is now bringing in a net annual profit ranging
between lire 1 20,000 and lire 1 80,000. The mean of these

* Samaritani, Prov. necessari nella laguna di Comacchio, p. 14.
f Bullo, loc. cit., p. 369. X Samaritani, loc. cit., p. 6.

57

figures or lire 150,000 (Rs. 90,000) is a fair approxima-
tion of the net average annual profit derived by the
Commune during any series of years marked by normal
conditions. Such a series is now current and a similar
cycle lasted from 1869 to 1891 as noted above when the
income varied between the same limits.

Samaritani arrives at identical results in a calculation
of the net annual profits upon the working of a period
of 15 years ending with 1883. For this series he states
the average yield of fish as : —

843,564 kilos of eels. 74>I24 kilos of mullet.

He calculates that two-thirds of the eels will be
distributed in the marinated condition at an average
price of lire no per quintal (Rs. 66 per 220 lb., say 4
annas 10 pies per lb.), and that the remainder of the
fish will be sold fresh, the eels at lire 80 per 100 kilos
(Rs. 48 per 220 lb.) and the mullet at lire 60 per 100
kilos (Rs. 36 per 200 lb. or 2 annas 8 pies per lb.).

On this basis he estimates the proceeds as follows :-

Two-thirds of the eels, sold marinated at lir
1 10 per quintal

One-third sold fresh at lire 80 per quintal

74,124 kilos of mullet at lire 60 per quintal

Sale of smelts and miscellaneous fish

Sale of hay and rent of pasturage

Shooting rents

Sundry receipts ...

Average gross annual receipts

LIRE.

6i8,6i3-6o
224,95°"4o

44,474'4°
6,ooo'oo

1 1,800*00
3,6oo-oo
5,ooo-oo

9I4»438'4o

The annual expenditure does not vary in correspond-
ence with the fluctuations in receipts, and may be taken
to average 750,000 lire, the precise mean of 26 years
being given by Samaritani as lire 755,230. Deducting
this sum from that given above as the average gross
receipts, we find the net annual profit of the undertaking
to be lire 159,208. The area of the lagoon during these
years was 39,500 hectares, so this sum is equivalent to a
working profit of lire 4-05 per hectare or lire r66
(exactly one rupee) per acre, an extremely unsatisfactory

§8

return compared with that at Arcachon which as stated
on page 17 amounts to as much as Rs. 60 per acre.

The significance of these figures led Samaritani in
1884 to advocate the relinquishment of fish farming in
favour of reclamation of the whole lagoon. He was then
Engineer to the Communal Council of Comacchio with
intimate knowledge of local agricultural and physical
conditions. In a report to the Council in the year
named he discussed in detail the advantages of such a
policy. The gist of his conclusions was that if his pro-
posals were carried into effect, instead of a present annual
profit of about lire 150,000 made from fish farming, the
commune may expect to obtain an annual revenue of
from lire 2,370,000 to lire 4,345,000 if the lagoon were
reclaimed for the cultivation of rice, beetroot, hemp and
grapes, the crops which yield best in this region. His
figures were as follows : —

LIRK.

i. Annual income calculated from the returns
actually obtained from sections of the lagoon
already reclaimed ... ... ... ... 4,345,003

2. The same calculated on the mean return from

all kinds of reclaimed land in the province of

Ferrara , ,,. ... ... 3>367>375

3. Revenue based on the returns from lands occa-

sionally flooded by the sea ... ... ... 2,370,000

These returns are equivalent to an annual revenue of
lire no, 85^, and 60 respectively per hectare, as com-
pared with lire 4 per hectare yielded by the lagoon as a
fish-farming establishment. Samaritani concluded by
stating that while the first rate is that which he believed
to be reasonable from a study of existing facts, the
second should be acceptable by the dubious and the
third should convince even those who are incredulous of
the great benefit to be derived from reclamation.

I ought to mention in this connection that the
conditions for profitable reclamation are exceptionally
favourable at Comacchio. The general level of the
bottom of the lagoon lies from i| to 2 feet above low
tide level and several series of embankments parallel
with the coast already exist. If the two seaward open-
ings be closed and the existing sea-canals converted
into main drainage channels, the expense of reclamation
would be comparatively small.

59

So far the commune has preferred to pursue a
middle course — to improve the conditions in the better
class of their fish farms and concurrently to undertake
the reclamation of the sections less favourable for fish
rearing, so that over 3,000 hectares of what was water
surface at the time of Coste's visit have now passed
permanently into agricultural occupation. The process
is likely to be continued until the whole inner half of the
lagoon, lying from 10 to 20 kilometres distant from the
sea, shall come under tillage, leaving the seaward half to
continue as fish farms.

Conclusions.

Comacchio furnishes to the fish culturist many object
lessons particularly valuable in India where the efficient
utilization of extensive tidal backwaters analogous to
that of Comacchio is one of the principal problems now
before the Madras Fishery Department. These demons-
trations fall into two divisions, those that are lessons in
methods that are admirable and those that are instances
of faulty and defective procedure to be carefully avoided
or materially modified and improved.

In this connection it is necessary to emphasize the
fact that the lagoon of Comacchio is in reality not an
establishment of fish culture properly so-called ; it is
merely a complex system of lagoon fishing whereby the
fish which enter as fry have their escape barred, to be
captured en masse as successive broods reach maturity,
a system handed down without appreciable alteration
through fifteen centuries. In the middle ages it appears
to have given much greater returns than the average for
the past two hundred years, and present methods call
for material modification if the yield of fish per acre is to
be increased to adequate proportions. At present the
yield does not average more than 50 lb. per hectare or
about 20 lb. per acre, a wholly inadequate rate consider-
ing the known potentialities of water areas cultivated on
scientific principles. Empirical as are the methods of the
Arcachon fish farmers, they actually produce per acre
twelve times the weight of fish yielded by Comacchio for
a similar area, the average of two years' yield in France
being 608J lb. per hectare annually.

6o

■

To formulate in detail any scheme for the improve-
ment of the Comacchio industry does not enter into the
scope of this report ; for present purposes it will suffice
if I enumerate what in my eyes appear to be the principal
defects of the existing- system and note the general
principles upon which any permanent improvements
must be based.

In the first place the present area is too extensive to
receive adequate cultural attention. Worse still, the
major portion is so remote from the sea that it is impos-
sible to provide means for tidal freshening, a measure
essential to the permanent well-being of a lagoon such
as that of Comacchio. Experiment has shown that in
the canals connecting the valli with the sea, no tidal rise
and fall is marked at a distance of 10 kilometres from
the sea mouth. The two great valli, Mezzano and
Fossidaporto, constituting almost one-half of the lagoon,
are both beyond this distance from the sea and hence
cannot possibly be benefited appreciably by ordinary
tidal flow. The frequent freshening rightly considered
to be vitally important in the case ol French fish farms
is virtually absent here. This lack of control over these
great expanses of water is the basal reason for the
disasters which punctuate the history of the lagoon, and
till it be removed the prosperity of the establishment is
at the mercy of the seasons. All other defects are of
very minor importance compared with this.

Were the fish farm area of the lagoon reduced by
means of the reclamation of the sections remote from
the sea, and effort concentrated upon a number of small
valli contiguous to the sea-bar, it would, I am positive,
be at once possible to render stable the commercial
success of the industry by eliminating or at least reduc-
ing to unimportance the dangers and losses hitherto
attendant upon the occurrence of abnormally dry seasons.
It would then be possible to devise means for the
admittance of tidal water at frequent intervals and in
large volume, the salinity of the enclosed waters would
never appreciably exceed that of the sea, large supplies
of plankton would be available as food, and in presence
of the healthier conditions set up by the alternate ebb
and flow of large volumes of water, the growth of
the fish would be accelerated. The great losses now

6i

experienced on account of wholesale poaching would be
materially lessened, the reduced area permitting of more
effective supervision. There can be little doubt that
under these improved conditions the yield of fish would
be largely increased ; the present yield per acre is so
small that a very slight improvement in methods
should enable double the present catch to be made — -
40 lb. per acre is all the quantity required in order to
make such improved seaward half of the lagoon yield as
much revenue, at less expense, than does the whole area
in its present condition. In addition, the Commune
would then be receiving rental upon 45,000 acres of agri-
cultural land obtained by reclamation of Valli Mezzano
and Fossidaporto. Under wise and energetic guidance
the Commune should be able to develop its lagoon into a
property yielding not ,£6,000 a year but £"50,000 at the
very least — reclamation of half the present area should
bring in not less than ,£45,000 (the rental calculated at
£"1 per acre) while if fish culture and not mere fishing
be practised in the remaining area, the yield therefrom
in fish should be quadrupled without difficulty in which
case net profit of ,£10,000 to £"12,000 should accrue
from this industry.

When I say fish culture instead of fishing, I mean
that instead of herding all species together as at present,
separation to some extent should take place, and that
special attention be given to the introduction and culture
of other species beyond those now reared. The fry of
some of these would probably have to be supplied from
hatcheries, but in these days when plaice, and cod and
shad are hatched by the million under artificial condi-
tions, this would present no difficulty. The provision of
artificial food to supplement natural resources in selected
ponds of limited area is another innovation likely to
prove economically sound ; it would enable the yield to
be enormously increased of those ponds where it might
be possible of introduction. But these are details that
must be decided and worked out by an expert on the
spot.

In spite of all imperfections, the present system has
served Comacchio well; it has been the mainstay of the
population for 1,500 years and the reluctance exhibited to
risk the consequences of any radical change need not be

62

a matter for surprise. Directly and indirectly the fishery
furnishes the means of regular subsistance to some 2,500
people and furnishes a surplus for the general purposes
of the city which seldom falls below ,£6,000 per annum.
Comacchio is a remote and isolated community and such
is ever intensely conservative and loth to embark upon
new enterprises so long as day to day needs are met under
the system handed down from preceding generations.

6

■>

III. — THE SCOPE FOR MARINE FISH-
FARMING IN INDIA.

In the foregoing pages have been described the fish-
farming methods practised in France and in Italy, the
only countries where this industry is carried on upon a
larore scale. We have now to consider to what extent
and with what modifications the principles underlying
these systems may be adapted to Indian conditions in
order to increase the food-supplies derivable from back-
waters, estuarine creeks, and sea-ponds along the coast.

This enquiry requires consideration under three
heads, namely : —

(a) The species of suitable fish available in the

Madras Presidency.

(b) The waters available and suitable ; their cha-

racters and extent.
(Y) The procedure considered most likely to prove
successful as being adapted to local condi-
tions.

The Species of Fishes suitable for the pur:poses of
Marine Fish-farming in Madras.

At Arcachon, the marine fish found to thrive best and
give satisfactory results in captivity are eels, mullet of
several species, and the bass (Labrax lupus). The fry of
soles and plaice also enter the ponds in large numbers
but as no means are taken to confine them in separate
enclosures all but a small number fall victims to the
rapacity of old eels while yet quite small owing to their
sluQ-o-ish habits and life preference for the bottom. Mullet
and bass, although they too pay a heavy toll to the eels,
survive to maturity in considerable numbers as they are
much more active and restless in their movements than
flat fish and when young pass most of their time swim-
ming near the surface in shallow parts of the ponds little
frequented by the mud-loving eel. The mullet like the
eel is tolerant of even considerable changes in the sali-
nity of water ; it feeds largely upon minute plant life
which is usually more abundant in brackish water than

64

in the open sea, and for these reasons and because it and
the active and omnivorous bass and the voracious eel
have different feeding habits, these fishes form an
admirable trio for the purpose of the fish farmer in
France.

At Comacchio eels and mullet are found to live well
together ; these two species form the bulk of the produce
of the lagoon. Sand smelts {Atherina boyeri) and
Gobies (G. lota), furnish considerable quantities in some
years, the former giving from 70,000 to 90,000 kilos per
year, the latter from 2,000 to 3,000 kilos. These two
fishes apart from the direct contribution they make to
the revenue of the fishery are of immense importance as
food for the eels — the immense shoals of aquadelles, as
Atherina is termed at Comacchio, constituting their chief
nourishment during the greater part of the year. The
other species of fish which thrive in these fish farms
comprise bass, soles, plaice and sea-bream (Spams
aura to) but although they constitute a notable propor-
tion of the fry which enter from the sea comparatively
few survive to maturity.

Shrimps {Crangon vulgaris) and Carcinus viacnas,
the common shore crab of Europe, enter in enormous
quantities when the sluices are open for the inontata in
the spring and as they also breed freely in the lagoon,
they furnish an indispensable and never failing source
of fish-food.

The ability not only to withstand and survive com-
paratively considerable variations in the salinity of the
waters they inhabit, but to thrive almost equally well
whether the salinity be high or low — within reasonable
limits — is the most important attribute to be considered
when deciding what species are best fitted for culture in
marine fish farms. The ponds employed cannot be put
in fully free communication with the sea, and hence in
dry seasons may tend by reason of rapid evaporation to
become comparatively highly saline, while on the other
hand the occasional torrential rains and high floods which
are factors in the situation in India wheresoever the
pond might be situated, are bound to reduce the salinity
below the normal at certain seasons.

Unfortunately there is lack of precise information
available on this point. Among European fishes suitable

65

for pond culture * it is a matter of common knowledge
that eels and certain species of mullet have been found by
experience to possess the widest range of tolerance, the
bass not far behind, and Bullo has compiled a table
giving the maximum and minimum range of tolerance of
salinity for these and other lagoon fish. The figures
given are admittedly tentative, the author stating that the
table is founded on deductions from rough data provided
by the empirical working of Italian fish-farms ; it cannot
be accepted except as an approximate guide. It is
probably unreliable in details, but since it is almost the
only tabulation of the kind we possess, it will be of use
to reproduce it here. The figures are as follows : —

Table showing the limits of salinity tolerated by the principal animals
reared in Italian lagoons, after Bullo. f

Name.

Limits of tolerance.

Minimum

Maximum

(parts per

(parts per

1,000).

1,000).

24

40

...

24

35

...

20

40

...

21

40

...

25

40

...

20

40

...

16

40

...

IO

40

...

IO

40

...

5

40

...

4

40

0

40

...

26

40

. . .

26

40

• • .

16

40

...

25

40

18

35

Fishes.
Goby {Gobius lota)
Golden mullet {Mugil auratus)
Smelt {Atherina Boyeri)
Sea-bream {S partis aura fa) ...
Sole {Solea vulgaris) ...
Plaice {Platessa passer)
Leaping mullet {Mugil salt ens)
Thick lipped mullet {Mugil chela) ...
Bass {Lab rax lupus)
Grey mullet {Mugil capita) ...
Large headed mullet {M. cephalus) ...
Eel {Anguilla vulgaris)

Crustaceans.

Common shore crab {Carcinus j/iaenas)
Common shrimp {Crangon vulgaris)

Molluscs.

Common cockle {C. edule) ...
Edible mussel {M. edulis) ...
Oyster {Oslrea edulis)

* Anadromous fish such as the salmon and the sturgeon are excluded for
obvious reasons.

f Loc.
5

cit. p. 176.

66

Considering that the salinity of the open sea in
Europe is about 35 per 1,000, this table credits none of
these fish with a tolerance of exceptional salinity. Its
chief inaccuracy appears to be in regard to this state-
ment ; later writers, particularly Samaritani and Feletti,
adduce facts which show that the eel in particular can
endure a much greater salinity than 40 per 1,000. The
former writer states that to his knowledge eels are able
to adapt themselves to a salinity of 8° Beaume (= 80
grams of total salts to 1,000 grams water) ; further that
in June, July and August 1893 the salinity of the
Comacchio valli, and in particular the Mezzano, rose
everywhere to 70 and 8° Beaume and in a certain locality
even to io°, without, in that year, causing any mortality.
He notes specifically that the fish were sana e vivacissima.
How far a high salinity, say between 40 and 60 per 1,000,
is inimical to ihegrowth of eels is unknown ; experiments
upon effects produced by variation in salinity upon
fishes in captivity are urgently needed — we require to
know the limit of tolerance wherein each species can
exist, and, more vitally important even than this, we
must obtain precise data as to the optimum of salinity,
in order that we may eventually regulate this to the
degree best suited to the rapid growth of the fishes
under control.

Eels, several species of mullet, smelts, bass, sea-bream
[Spar us an rata), gobies, plaice, and soles being the
chief food fishes which are found to thrive in captivity
in Europe, it is satisfactory to know that fishes of closely
related species and of similar habits are found in Indian
waters in fully as great numerical abundance as in
Europe. Eels are represented by two species, mullet by
not less than 27 species, several being noted for their
marked partiality for backwaters, and some being actually
domiciled their whole life in rivers, smelts by three
species of Atherina ; in place of the bass we have the
finer Kodawa {Lates calcarifer) ; sea-bream are repre-
sented by several species of Chrysophrys which resort
habitually to backwaters at certain seasons ; gobies by
numerous species of the same genus ; plaice by the fine
Pseudorhombus arsius, and soles by several species of
estuarine-loving Cynoglossus. The fry of certain species
of all these oenera of Indian fishes resort to backwaters

o

6;

and estuaries in vast multitudes at certain seasons, and
from the fact that their European representatives thrive
in captivity, these kinds are marked out as specially
well adapted for cultural purposes. But this list does
not exhaust by any means the number of kinds which
appear suitable for culture ; it omits several of the very
best, species probably of greater value than the majority
of those already named. Principal among these addi-
tional species are the various kinds of bamfn or kala
(Polynemidae), and the so-called " whiting " of Madras
(Sillago spp.) ; the fry of all these frequent estuaries
and, according to Day, so also do the young of that most
excellent of fish, the white pomfret {Stromateus sinensis).

At the present time we know almost nothing definite
about the habits of Indian sea-fishes ; the acquirement of
precise information on this subject in regard to the more
esteemed of our food fishes is pressing, and the selection
of fishes for the initial culture experiments must be
tentative. We have yet to find what fishes will thrive
best under the peculiar conditions attendant upon con-
finement ; we may easily make a wrong choice ; we have
also to find what species may be reared in the same
ponds, and we have to learn whether we can depend
upon the natural montata or " mounting " of the fry to
supply the ponds or whether it will be preferable to
hatch fry artificially.

Two distinct considerations enter into the problem
of the selection of species for cultural experiment — the
first is hardiness to bear occasional variations in salinity,
to endure prolonged confinement, and a diet more
restricted perhaps than that normal to the species, the
other is the quality of the fish from a table point of view.
We have already seen that eels and mullet belong to the
first category, but we are ignorant of how far the bamfn,
the " whiting " and the white pomfret conform to this re-
quirement, and it happens that these latter fishes together
with the mullet, are the kinds most esteemed of the whole
number available, whereas eels and catfish, hardiest
probably of all, are those of least commercial value in
India, whatever the former may be for purposes of
export.

A few notes detailing the salient facts of cultural
import concerning the species enumerated may usefully

68

be given here, but they are at the best meagre and their
amplification should be one of the first and most import-
ant tasks to which the attention of Indian fishery officers
should be directed. Efficient work is impossible without
satisfactory and sufficient data and these we do not
possess.

Eels. — Of these two species are recorded by Day *
Anguilla bicolor and A. bengalensis. These fish, the
veldngu and pdmbu min of Tamil districts, abound in
muddy creeks and estuaries on both the East and West
coasts. They are not often caught by fishermen and
then only by accident, as no native will fish for them
purposely owing to the very general objection of natives
of India to eat them owing to their snake-like form and
supposed absence of scales — the latter being specially
objectionable in the eyes of Muhammadans. The size
attained by eels in India appears fully as great as in
Europe and there can be little or no doubt that their
life-cycle and habits coincide with those of related species
in other parts of the world.

The remarkable migration of young eels or elvers
from the sea into rivers and fresh-water lakes, their
sojourn there for 4, 5 and even 6 years, and their eventual
return to the sea to breed, are phenomena known for
centuries and utilized commercially as we have seen by
Italian fish farmers for at least 2,000 years. Where and
under what conditions the sexually mature eels congregate
and breed and what the larvae are like remained an
enigma until a few years ago. The researches of Grassi
and Calandruccio first solved the latter problem ; they
were able to demonstrate by direct experiment and other-
wise that certain forms of pelagic ribbon-shaped fish-like
creatures, previously known as Leptocephalids and con-
sidered to belong to a special and separate family, were
in reality the larvae of the common European eel. The
species of Leptocephalus so identified is the one named
L. brevirostris, Kaup. Eingenmann and Kennedy sub-
sequently established the identity of another from, L.
Grassiiy Eing. & Kenn., as the larva of the American
eel, Anguilla chrysypa. Leptocephalids have been found
by both Day and Thurston in Indian seas ; the latter

* Fauna of British India — Fishes, London, 18S9.

69

observer in Bulletin No. 3 of the Madras Government
Museum (1895), mentions that he ''obtained a few
specimens in the Gulf of Manaar and a large number
from the meshes of fishermen's nets at Gopalpur, where
they are known as sea-leeches." * At that date their iden-
tity with the eel-family was unknown and Dr. Thurston
quoted Gunther's strange theory that they represented
the larvae of various marine fishes living a pelagic life
in an arrested and abnormal stage of development !

The identity of these larval fishes once established,
the completion of our knowledge of the life-history of the
eel became a question of the greatest interest to marine
zoologists and to-day many of the blanks have been
filled up. The range of Leptocephalid larvae was ex-
tended into the Atlantic by the capture by Schmidt and
also by Holt in 1904 of single specimens in the Atlantic.
The following year Schmidt obtained numerous indi-
viduals by means of the Petersen net, and was thus
enabled to determine some of the oceanographic condi-
tions under which they occur. The chief facts ascertained
during these memorable investigations are that the
leptocephalids of North European eels are not hatched
in shallow seas and that they are most numerous along
the course of the narrow band lying between the 500-
and the 600-fathom line of the Eastern Atlantic. They
were taken at various depths down to 1,000 metres, a
depth where the temperature is above 420 Fahr. and
salinity above 35 per 1,000. The greatest hauls were
made in the neighbourhood of a depth of about 70
fathoms. Other facts lead to the belief that the breed-
ing of the eel occurs at great depths. Large specimens
of the adult form of the eel have been obtained by Grassi
and Calandruccio from the whirlpools in the Straits of
Messina which occasionally bring to the surface animals
known to live at abyssal depths ; these eels, while having
the dark livery of the breeding season, were more deeply
pigmented and were marked by an extraordinary increase
in the size of the eyes. Vaillant also records the
extraction of an eel, 90 centimetres long, from the
stomach of a Cachalot whale captured off the Azores.

* The latter probably were not true Leptocephali ; it is more likely they were
elvers. J. H.

These and other observations, taken together with the
above-mentioned facts regarding the distribution of
leptocephalids lead us to conclude that the breeding-
place of eels is in depths somewhat greater than 500
fathoms in conditions where the temperature is above
420 Fahr. and the salinity above 35 per 1,000.

This brief resume of the state of our knowledge
concerning the early history of the eel shows the impossi-
bility of utilizing methods of artificial propagation in
the case of this eel. To stock fish-ponds with eel-fry,
dependence must be placed upon the 'migration shore-
wards of the shoals of elvers, the colourless fry into
which the leptocephalid larvae change when a certain
age has been attained. These elvers when they first
arrive on the coast average 7 centimetres in length,
actually smaller than the leptocephalid larvae from which
they come !

Eels in the sedentary stage are carnivorous and
extremely voracious. They are most inimical to young-
fish of all species, but particularly to such as live near or
upon the bottom. Wherever fish of greater value than
eels are to be reared, it is necessary to keep them in
separate ponds. At Comacchio, smelts and crabs form
the chief food of captive eels, a diet which should be
duplicated without difficulty in Indian fish farms.

Because of its rapid growth due largely to its indis-
criminate voracity and its ability to withstand great
variations of salinity, the eel is the most easily reared of
any species available for fish-farming in saline waters.
Precise data upon the rate of growth under definite
conditions are not available even in Europe, and, if they
were, it is probable that they would not be reliable guides
to that attained in Indian waters, where conditions are
usually more favourable. It may be mentioned however
that Coste, who was almost the first to give scientific
attention to the growth of fishes, stated that an eel fry
attains a weight of 2§ kilograms " within 4 or 5 years,"
while Gobin is responsible for the statement that "a kilo
oi mdntee oi eel fry containing 3,600 individuals, attain
in three years a weight of 6,000 kilos and a value of
3,000 to 3,500 francs (Rs. 1,800 to Rs. 2,100).*

* La Pisciculture en eaux salees, p. 212, Paris, 189 1.

7i

In Germany the rearing of eels in fresh-water ponds
is a wide- spread minor and accessory industry, and is
undoubtedly exceedingly profitable, the owners usually
marinating their products for local consumption.

In Northern Europe Gemzoe * has found from a
study of the concentric zones marked on the scales that
the age at which development of the breeding instinct
leads female eels to begin a migration to the sea ranges
from 6\ to 8J years, whereas in the male the migratory
instinct appears to be developed between 4 and 7J years
of age. It should be noted however that three years must
be added to the number indicated by the zonal markings,
as scales do not appear till two years of sedentary life
have elapsed, while if the deductions of Schmidt be
correct a third year must be added to cover the pelagic
or larval period. Hence the actual age of an eel is
three years greater than that of or indicated by the
scale. (In deducing age from examination of scale
markings, it is necessary to select those close to the
lateral line ; these are earliest to appear.)

The eel is a fish universally esteemed in Europe ; in
England although it is less popular than on the
Continent yet the local supply falls so far short of the
demand that London receives annually large consign-
ments from Holland.

Marinated eels as prepared at Comacchio and
throughout Germany form one of the most nutritious
and appetizing of preserved foods. Owing to prejudice
it would have a restricted consumption in India at the
commencement, but it has only to be once tasted to be
appreciated and, as a food and a luxury for the well-to-do,
it should attain popularity in spite of prejudice. How-
ever if a large and regular supply can be depended on,
local demand is not essential to success. If barrelled
and tinned as at Comacchio, an export trade to Australia
and Central Europe should be profitable as there the
demand is normally in excess of the supply — a demand
that increases annually in pace with the fecundity of
the German race, whereas the supply lacks elasticity and
corresponding increase.

* Age and Rate of Growth of the eel ; Report of the Danish Biological
Station to the Board of Agriculture, XIV, 1906, Copenhagen, 1908.

6-A

72

Mullet. — Mullet in European fish-culture rank next
after eels ; India is specially rich in species of this family
of fish ; they constitute the most abundant fish found in
estuaries and backwaters. In India the rano-e of habitat
is greater even than in Europe. Some species remain
permanently in rivers, other keep rather strictly to the
sea, while the remainder, which specially concern us,
enter backwaters and estuaries in immense numbers
both as fry and as adults. The most ingenious of the
nets used in backwaters, as at Ennore and Pulicat, are
designed to capture these wiliest of fish.

In confinement they exhibit in a large degree the
quality of hardiness ; they are easy to feed and will live
where most other fishes would die from starvation.
They feed largely upon vegetable growth and debris ;
at the same time they will feed voraciously upon shoals
of small crustaceans when opportunity offers. The
general opinion is that mullet with the exception of the
species domiciled permanently in fresh waters, spawn in
the sea. This I have reason to believe is not the case
universally * as I have seen fully grown female mullet
ranoqn^ from 18 to 21 inches lon^ taken from Ennore
backwater in numbers in the middle of January with
fully developed ripe roes. They were always accom-
panied by a larger number of males also in ripe condition.
As the females were of quite different dimensions to
those living in the backwater a month earlier, there is no
doubt these fish had entered the backwater from the sea
not long previously. A significant fact was that the
stomachs of the females were empty ; the stomachs of
the males were crammed with a tangled mass of green
algae. The former had apparently not entered the back-
water to feed ; the inference is strong ; they had entered
with intent to spawn. However this may be, it is a
matter urgently demanding settlement ; there would be
no difficulty in collecting an ample supply of ripe breeders
in January at Ennore and Pulicat if it were desired to
strip them when ripe and practise artificial fertilization
and hatching, procedure which is altogether preferable
to depending upon supplies of natural fry whenever a
skilled staff is available ; the former svstem allows full

* In Europe instances are on record of mullet thriving and breeding freely
in ponds fed principally by fresh water during the greater portion of the year.

73

control of the species treated and obviates the mixing of
different species of fishes in the ponds.

Mullet are industrious feeders ; apart from the ques-
tion of the possible direct absorption of carbon compounds
dissolved in sea water, they grow fat on a diet of mud
when this is laden with diatoms or other minute algae —
possibly even when such are absent and the organic
matter present has resulted from decomposition changes.
Thus these fishes pick up food where other species
would starve. They increase rapidly in size ; according
to Coste, mullet living in Comacchio lagoon increase
from 3 centimetres in length to 26|- centimetres (io|-
inches) during a period of 12 months. The table of
growth which he gives may usefully be reproduced here,
in order that a comparison with the rate of increase in
Indian waters may be instituted when this information
be obtained. Coste's figures are as follows : —

Age in months

I

2

3

5

6 7

9

10

12

Total length in centi-
metres.

n

j

5

8-5

I3-S

16-5 19

21

24

26-5

Greatest girth in centi-
metres.

1 '3

2-0

4-2

7-0

8-o 9-5

II'O

12*0

13-5

From my knowledge of the life conditions in Indian
backwaters, I feel assured that these figures are con-
siderably exceeded here ; indeed I expect to find that
mullet attain a oood marketable size at the end of one
year from the time they enter the ponds as fry.

Polynemids, the kala of Tamil fishermen, and the
bamin of anglers, constitute another family of food fishes
in high esteem in India. Off Point Calimere and in Palk
Strait they are the subject of an important special fishery
and along both the East and West Coast they are eagerly
sought both by professional fishermen and by the votaries
of Isaak Walton. Either fresh or salt cured, bamin is
one of the very best flavoured fish for the table, and Sir
Frederick Nicholson has found it to give an extremely
satisfactory product when smoked. So highly esteemed
is it that it is called sometimes the Indian salmon.

P. tetradactylus is probably the most valuable
species on the Madras Coast. It is the largest among
Indian species, attaining 6 feet and upwards in length

74

when fully grown ; it is tolerant of changes in salinity and
exhibits great partiality for backwaters and estuaries.
For these reasons it is indicated as a species well adapted
for rearing in confinement. The larger polynemids are
essentially predatory fishes. They are the special and
particular enemies of mullet, following them into back-
waters and penetrating far up tidal rivers in the pursuit.
Some species enter rivers and backwaters for spawning
purposes ; as with the mullets, there should be no diffi-
culty in obtaining fertilised ova for artificial hatching.

Provided sufficient supplies of suitable food be present
or procurable, polynemids should be found to be among
the three most profitable kinds of fish available in India
for fish-farming. They grow rapidly, attain a large size,
are tolerant of the variability characterizing backwater
life-conditions, and are obtainable as fry in the requisite
abundance ; finally, when put upon the market they will
always command a ready sale at prime fish prices whether
fresh or smoked.

The White pomfret {Stromateus sinensis) is another
of the highly valued Indian food fishes which it may be
found possible to utilize for pond culture. The young
are said to frequent estuaries and it is probable that they
would bear confinement well. They feed largely upon
small crustaceans, worms and the like, and their food
supply should not present any serious difficulties. At
present their breeding habits are little known and we
cannot go further than express the opinion that the
probabilities are that their eggs should be readily obtain-
able for artificial hatching purposes and that the fry are
likely to thrive well in backwater ponds. Experiments
must be undertaken to ascertain the most suitable
conditions under which successful culture on a large scale
may become possible.

Were the present Madras Aquarium capable of
accommodating a small fish-hatchery consisting either of
batteries of Macdonald jars or Dannevig oscillating
boxes, the necessary facilities for this and other investi-
gations would be available at a very low cost ; unfortu-
nately when the building was designed no provision for
this purpose was included and it will now be necessary
to build a separate hatchery with observation tanks, etc.,
for the conduct of fishery experiments.

75

Mullet and pomfret should go well together in the
same ponds ; in feeding habits they appear to be comple-
mental, the one fond of plant growth, the other of the
small animal life easily propagated in ponds. Neither
would interfere inimically with the other. What are the
prospects of acclimatising in ponds the fry of the other
two species of pomfret (the grey and the black) cannot be
stated, so little is known of their habits. In favour of
the possibility of so treating them is to be counted their
somewhat phlegmatic habits. They are comparatively
slow in their movements and fish of this temperament
usually adapt themselves readily to the irksomeness of
confinement.

The Sea-breams {Sparidac) are well represented in
India, but only those of the genus Chrysophrys appear
suitable for pond-culture. The different species run
generally from a foot to 18 inches in length. Shoals of
their young are frequently met with in backwaters at the
beginning of the south-west monsoon and they have the
same habit of working up estuaries against the current as
is characteristic of European species. They make excel-
lent eating and might prove useful fish for pond-rearing.

The Koduva (Lates calcarifer) is a large fish highly
valued, and well known to Europeans in Calcutta under
the names of begh and Cock-up. It is one of largest of
Eastern sea-perches and is almost the only species which
frequents backwaters and tidal rivers, so agreeing in
habit with its European relative the bass (Z. lupus). In
Pulicat Lake it abounds, particularly in the inner or
northern section, where the fishermen of adjoining
hamlets co-operate periodically in extensive driving
operations using cotton rope nets of great strength and
large mesh, the fish often ranging between 2J to 3 feet,
the maximum recorded being 5 feet with a weight of
200 lb. The air-bladders (sounds) of these fish are
valuable for isinglass manufacture, so this fish well
deserves attention for fish-cultural purposes. Its voracity
may prove an objection, unless some cheap source of
suitable artificial food can be provided. Otherwise it
should thrive well in captivity.

The so-called " whiting" of Madras, Sillago sihama
the karangan of the Tamils, is yet another excellent food-
fish, particularly well adapted to pond-culture. This and

76

a second species, S. panijns, readily adapt themselves to
brackish water ; they are fond of ascending tidal rivers
and of living in backwaters. They are compact little fish
seldom reaching a foot in length but esteemed amongst
the most wholesome of the fish brought into Madras
market. Special attention should be given to these fish
in any cultural experiments that may be attempted in
backwater ponds.

Smelts. — As food for carnivorous fishes such as eels
and polynemids, smelts will occupy an almost indispen-
sable position in the economy of Indian fish-farms. They
belong to the genus Atherina of which several species
are found in abundance in Indian waters. Like their
congeners in Europe, they affect backwaters and estuaries
in shoals of immense numbers during the fry stage, which
occurs during the cold season. Thus on Christmas day
1 90S, the Ennore backwater was literally alive with the
fry of Atherina forskali ranging in size from f to
1 inch in length. Children were catching them in large
quantities by the simple expedient of sieving water
through their cloths. A push- net used in the shallows
at this season would supply ponds of polynemids with
enough food to last them for months.

Flatfish, — Certain of the Indian species of flatfishes
are capable of thriving in brackish water, notably several
of the tongue soles, Cynoglossus spp. and the plaice-like
Pseiidorhombus arsius. These are prime table fish, and
as soon as a preliminary investigation shall have made us
acquainted with their breeding habits and the nature of
their habitual food, we shall be in a position to rear them
in quantities to adult size. Immature individuals are
caught at times in immense numbers in many of the
West Coast backwaters, and these catches should prove
invaluable to enable ponds to be stocked with fish which
have already emerged successfully from the myriad
dangers that beset them as fry.

Catfishes, especially of the genera Arius and Mac-
rones > would live and thrive well in mud-bottomed ponds,
but their market value is so trivial that it would not pay
to rear them in confinement, indeed vigilance will have to
be exercised to ensure their exclusion.

The list above given by no means exhausts the
number of species which appear capable of living and

77

thriving in confined areas liable to become more or less
brackish during the rains. These other species are
however of minor importance, either in little esteem for
the table or too small to be useful except as food for
bamin, eels or begti.

The number of really useful species which are avail-
able, namely, bamin or kala, mullet, kodava, sea-bream,
Sillago (Madras whiting), smelts, plaice, soles and eels,
include all that are considered of importance either in
the Arcachon fish-ponds or in Comacchio lagoon, and
there is reason to believe that the bamin and the kodava,
both of greater value from the Indian standpoint than
either mullet or eels, will prove susceptible of propagation
and rearing in enclosed areas. Neither of these species
is available in Europe, so marine pisciculture in India
has the initial advantage over Europe of having both
better quality and a greater number of suitable fishes
wherewith to experiment. The rate of growth in India
is probably much more rapid, but upon this no data are
at present available.

The Waters available and suitable for Marine Fish
Culture in Madras ; their Characters and Extent.

The places susceptible of conversion into fish farms
fall under three heads, namely : —

(a) Backwaters and estuarine creeks.

(b) Salt pan channels.

(c) Estuarine or deltaic marshes.

(a) The first category includes all the large areas of
landlocked tidal water wherein fish rearing can be
carried on. I say advisedly rearing and not culture, for
in the case of ponds having an area of more than a few
acres, a modification of the natural culture as practised at
Arcachon and Comacchio must necessarily be the system
pursued in contradistinction to that intensive or artificial
culture whereby fish are reared from artificially impreg-
nated ova and fed partly upon artificial food, a system
applicable only to ponds of comparatively small area.
Both the east and west coasts of the Madras Presidency
abound in backwaters, many of them veritable inland
seas. On the East Coast, Pulicat lake is the largest of
these tidal lakes, having a waterspread of 178 square

73

miles in the rainy season. Covelong and Dugaraza-
pattanam are two others of large size well designed
by nature for conversion in part into fish farms that
might easily surpass Comacchio, were capital and orga-
nisation available. The backwaters on the West Coast
are even more extensive than these two latter ; they are
richly supplied with branches and side-creeks which
swarm with myriads of fry at the commencement of the
rains and which require little else than bunding off from
the main channel to convert them into natural fish-ponds
of any required area.

The side creeks, of all manner of size and aspect,
given off in great numbers by all the great rivers in the
neighbourhood of their embouchures are equally suitable
for natural fish culture on a large scale wherever they
can be cut off securely by bunds from overflow by flood
water. In the deltas of many of the great rivers on the
East Coast there are often to be seen long stretches
of quiet water representing channels forsaken by the
river for some new outlet ; these old channels are often
ideal situations for fish rearing as they are well within
tidal range and would also afford frequent opportunity
for freshening with sweet water when this became
desirable.

The presidency is particularly favourably situated
for the inception of large schemes for the natural culti-
vation of sea fish on the Venetian system. So many
places are so perfectly adapted for it, that it is difficult
to point to one as being notably superior ; all circum-
stances considered I am inclined to think that whereas
Dugarazapattanam backwater may be equalled, it can-
not readily be excelled as the site of a lagoon fish farm.
The area available is ample in acreage, the depth is of the
requisite shallowness, a sufficiency of fresh water is
available during the rains to reduce salinity and induce
a strong run of fry whereof multitudes belonging to
suitable species swarm in from the sea on the annual
opening of the bar ; finally, prescriptive fishing rights
are non-existent in the backwater, the fishing popula-
tion being most meagre, their operations virtually
limited to the open sea. In such circumstances the
initiation of a fish-farming industry at this centre would
come as a boon to a population living under conditions

79

of marked poverty — the prosperity of coastal villages in
this district having- suffered considerable diminution
through the competition of the East Coast railway with
the coasting trade and with the Buckingham Canal,
formerly the highway of traffic up and down the coast.

(d) Salt-pan channels.

One of the most promising openings for sea-fish
culture, properly so-called, appears to lie in the utilisa-
tion of the salt water feeder channels at certain of the
many Government salt factories dotted along the East
Coast. The channels in question usually connect by
means of well-built sluices with neighbouring tidal
creeks ; their function is to convey sea water to all
sections of the salt pans within each factory. Very
commonly the channel makes the circuit of the pans
which it feeds by the intermediary of smaller channels
at a higher level ; this main channel has the form and
appearance of a moat surrounding the salt pans. On
the outer margin through a great part of its length it is
bounded by a substantial embankment rising usually
to 2 feet or more above the highest flood level of the
adjoining creek and considerably more than this height
above high water of spring tides. The principal channel
is maintained several feet in depth, nominally 6 feet,
I understand. Somewhere in its course the channel
usually widens into a pond-like expansion of varying
depth or there may be several of these.

From the channel and its branch ponds water is
raised by various kinds of lifts into the high level
gutters. It is used solely as a canal to furnish the
supplies of salt water required for the manufacture of
brine ; provided this function be maintained it does not
matter to the salt lessees what secondary use be made
of the channel. Fish-culture in the main waterway of
many factories can undoubtedly be carried on without
detriment to the manufacture of salt ; it may, where
considered necessary, be restricted to particular sections,
permeable barriers being placed where needful to pre-
vent fish entering the sections where water is being
taken out for brine making. What the salt makers want
is an ample supply of clean saline water ; in fish-culture
the same need prevails, hence it follows that being

8o

equally necessary for both industries, there could arise
no disagreement, while in the event of fish-culture being-
carried on concurrently with salt making, the work of
the fish-tenders being directed to the maintenance of the
water and the channel in a clean and wholesome condi-
tion, the water supply for the pans would, if anything,
be benefited. It is a matter of common knowledge that
the presence of fish in a pond has a direct influence for
good, the water undergoing distinct purification through
the removal of matter that would otherwise putrify and
tend to contaminate it.

However no decision need be come to on this
question for some time and by then we shall be in
possession of concrete data, for there exist some salt
factories where the feeder channel although present, is
in disuse, water for the pans being obtained in pits by
percolation. Here then are no possible difficulties in
the way of instituting a working experiment and if that
should prove successful and it be found, as I am sure it
will, that such operations in no way interfere with the
regular routine of the factory and actually have a bene-
ficial effect through rendering the water supply cleaner
and more wholesome, then the Madras Government may
utilize the channels of many of its salt factories for a
remunerative secondary industry at almost no extra cost,
seeing that nearly all the machinery for such work is
already in existence in the shape of sluices, channels,
and protective embankments.

(e) Estuarine and deltaic marshes.

On both coasts there are hundreds of square miles
of low-lying ground bordering the lower reaches of our
rivers which are overflowed whenever the adjoining
stream comes down in flood or when high spring tides
invade the estuary. It has been from land of this nature
that salt pans have been formed and in the same way it
would be a matter of little difficulty to make fish ponds
of whatever area might be required.

Even as it is the marshy land adjoining many streams
on the Malabar coast is the scene of a crude and undeve-
loped attempt at fish rearing, an attempt containing the
germ of true fish-farming. The land in question is very
low and is cultivated in plots at the time of the rains by

8i

the soil in places being heaped up in raised beds, small
plots rising from the excavated or lower level. With
the advent of the wet season the latter areas become
flooded and paddy is sown upon the raised plots which
rise like islands above the surface of the flood.
Concurrently a low bund is raised to enclose the general
area dealt with in order to retain a supply of water as
the rivers soon fall too low to maintain a proper inunda-
tion of the paddy planted plots. In the shallow stretches
of water thus retained by bunding, fish fry which came
up the river during the height of the flood are imprisoned
in countless myriads. After the paddy is harvested, the
water is still kept bunded to retain the little fish, which
are collected and sold as soon as the water begins to dry
up. Such or similar lands require but a limited expen-
diture to convert them into excellent fish farms for the
intensive culture of mullet, bamin, and eels if properly
selected within easy reach of a tidal supply of salt
water.

The area of marshy estuarine land of this description
is practically unlimited ; much of it is too saline for
agricultural purposes and lies waste and unprofitable at
the present time.

Procedure most likely to trove successful under
Local Conditions in India.

As already mentioned two distinct systems of fish
culture are indicated as applicable to Indian saline
waters, (a) the natural, as practised at Arcachon and
Comacchio and (b) the intensive which hitherto has
never been carried out commercially except to a limited
extent in Japan where it is confined to mullet and eels.
Each system has its merits under definite conditions and
each is susceptible of many modifications dependent
upon the varying nature of local conditions at different
parts of the coast, upon the species of fish to be reared,
upon the acreage available, the financial resources
disposable, and in less degree upon a number of other
considerations peculiar to each location and its neigh-
bourhood. Each culture pond and each natural fish farm
must be designed independently with due regard to
these various influencing conditions ; the environment
of each establishment must receive full consideration.

82

Much of these necessary data is not yet available ;
observations are urgently needed upon the breeding
habits and the nature of the food of the principal back-
water and estuarine fishes and experiments in artificial
impregnation and hatching are needed before marine
fish-culture can be carried on commercially with fair
assurance of eventual success.

Most useful information will become available if a
well considered experiment in fish rearing be carried out
by this department's officers in such a series of ponds as
are available at trivial expense by the conversion of
disused salt channels at several selected localities alongf
the East Coast. Detailed proposals to this end will be
prepared and presented separately. It is of interest
to note that the Arcachon fish farms were begun in
this very way — by conversion of salt-pan channels into
ponds.

Comparison of local waters and conditions so far as
are known, with those of Europe holds out definite
promise of success if plans be well drawn up and the
work judiciously managed. Several of the most impor-
tant conditions in India are greatly more favourable
than in Europe ; we have more and better species of fish
available from which to select the kinds which will repay
culture most profitably and the rate of the growth is, we
believe, much more rapid than in temperate latitudes *
owing to the great abundance of food in inshore and
estuarine waters and the absence of inclement winter
conditions when growth suffers arrestment in Europe.

Natural fish farming, based upon the rearing of fish
from fry hatched naturally in the sea involves so little
expense that under these specially favourable conditions
commercial success should be easv of attainment. Once
the pond be made and the sluices fitted — and these exist
already if we utilize salt-pan channels — almost the only
expenses involved consist of the wages of a sluice man
and a watchman. The fish under this system forage in
the pond for the whole of their food and receive or
require practically no attention save in regard to peri-

* The data obtained from our oyster-culture experiments at Ennore are
most definite in regard to this shellfish ; we have learned that eighteen months
from spawning are sufficient to provide well-grown oysters fit for market, ranging
in size between 3J x 3 inches and 4I x 3! inches.

83

odical freshening of the water and the throwing into the

• i • •

pond of any live food caught in the sluice passage during

this operation. The disadvantage of this system is that
control over the species admitted and reared is far from
complete. It is a matter of the greatest difficulty to
keep different species separate — indeed under Arcachon
and Comacchio methods all comers are welcome and all
are herded together indiscriminately. As a consequence
the hardy and voracious eel flourishes at the expense of
the other fish, the mullet coming next. Now in India
there is at present no demand for the eel. Except if
prepared for export this fish would not pay to be grown ;
indeed unless the fishes more valued locally can be
reared separately, the presence of large numbers of eels
in culture ponds would prove detrimental and might
easily prove fatal to the commercial success of the
operations. Intensive culture of fishes which command
a high price in local markets is more likely to pay better
than natural culture provided such fish can be put upon
the market in prime condition. If it be found possible
to rear pomfret and polynemids in ponds without exces-
sive trouble or expense, the esteem in which these fish
are universally held will determine these species as the
most profitable for intensive culture. Separate ponds
would probably be required for these two fishes, mullet
being added to the pomfret pond. But we must first
find whether the last named fish is susceptible of culture
in confined areas. In true intensive culture, arrange-
ments would be necessary to supplement the natural
food supply of the ponds by means of artificial food made
from slaughter-house refuse, fish waste and fish meal
obtained by grinding up excess catches of sardines and
catfish which otherwise would be utilised for no other
purpose except manure. To ascertain how far the
comparatively low rates obtainable for fish in this country
will permit of this phase of intensive culture is one of
the primary tasks to be undertaken in any practical
experiments in marine fish-farming.

TUTICORIN,

2jth November 1910. JAMES HORNELL.

MADRAS FISHERIES, BULLETIN No. 6.

Plate V

frfttMi^" "1

jmn|£j|piM*l'>"

EFrt 1 f«

■

Fig. 7. — Fish Labyrinth, Serilla Fishing Station, Comacchio.
View from the Covola looking toward apex of first set

of palisades.

Fig. 8. — The same labyrinth showing a lateral otele
wherein the eels are finally trapped.

[ /. Homell, Photo. ]

MADRAS FISHERIES, BULLETIN No. 6.

Plate VI.

Fig. g. — View of the terminal traps (oteli) ok two twin labyrinths,
Serilla Fishing Station, Comacchio.

Fig. io. — View from within the Cogolara of one ok the same
labyrinths, looking towards the mouth <>k the terminal

OT EI E.

|/. Hornell, Photo.]

MADRAS FISHERIES, BULLETIN No. 6.

Plate VII.

Fig. ii. — View from Serilla Fishing Station towards its sea-canal

(canal ungola).

Fig. 12. — A lock on a service-passage leading from a sea-canal
into Valle Cona, Comacchio.

[/. Horn ell, Photo.

Y

H IflXY i

L

SL(.+0