ON THE ACTION OF SENECIO ALKALOIDS AND THE

CAUSATION OF THE HEPATIC CIRRHOSIS OF

CATTLE (PICTOU, MOLTENO, OR WINTON DISEASE)

ARTHUR R. CUSHNY

Pharmacological Laboratory, University College, London

Received for publication, June 10, 1911

In a number of the colonies a disease affecting horses and

cattle and inducing hepatic cirrhosis has been known for some
years. In New Zealand, in which it occurs chiefly but not exclu-
sively in Southland, it is known as Winton disease, in South
Africa as Molteno disease, and in Canada as Pictou disease, from
its occurrence in the neighborhood of Pictou, Nova Scotia, ex-
clusively. The symptoms of the disease are described as prac-
tically identical in the three localities. Cattle’ are observed to be
“unthrifty” for some time, but definite symptoms appear only
three or four days before death, and commence often in diarrhcea,
diminished milk, dry and staring coat and disinclination to feed.
The diarrhcea is not necessarily severe, but is often followed by
straining which increases in intensity and frequency and may
lead to eversion of the rectum and rupture of it+s vessels. Con-
siderable pain appears to be felt, the cattle groaning and lying
down or becoming frenzied and charging anyone who approaches.
Eventually unconsciousness sets in and death follows in two to
four days after the first definite symptoms were observed.
The liver is found in some cases to present the appearance of
chronic cirrhosis and feels leathery and tough when cut. In
others the only change in this organ is marked venous conges-
tion. It is generally of small size and blue-slate in color and the
edges are rounder than usual. The gall-bladder is distended

1 W. H. Chase: Agricultural Journal, Cape of Good Hope, xxv, p. 675, 1904.

531
532 ARThUR R. CUSHNY

with very viscid yellowish black or black bile, and the interior
shows a number of red spots of the size of a pin’s head. The
urinary bladder and heart may also show these petechia. The
first three stomachs present nothing abnormal, but the fourth has
red hamorrhagic spots. The folds are thickened by the submu-
cous exudation of gelatinous fluid. The intestine is inflamed
around the openings of the bile ducts and the lower bowel may be
congested from straining.
The disease is of great economic importance, for in Nova
Scotia it is calculated it has been the cause of the loss of several
thousand head of stock, and in the East London district of South
Africa, Dixon2 states that it has rendered horse-breeding impossi-
ble as the mares contract the disease after two years or more
grazing. A number of investigations have been carried out as
to its causation more especially in Canada where its limitation
to the district immediately around the town of Pictou has been
especially remarked. It seems to be generally believed in this
neighborhood that the disease is of recent origin and it has been
associated popularly with the introduction of Senecio jacobaa or
ragwort (Stinking Willie) in the ballast of a ship from Scotland
about fif1y years ago. The disease certainly rose a few years
afterwards and occurred more particularly on the farms on which
the plant was prevalent. But investigations carried out in 1882
under government auspices failed to connect the disease with the
weed, and cirrhosis of the liver was therefore put on the list of
contagious diseases, animals affected with it were slaughtered
and the buildings in which they had been kept were disinfected.
The pathology of the disease was studied by Osler, Wyatt
Johnston and especially exhaustively by Adami,3 who states that
the main lesion is an extreme ondition of cirrhosis of the liver,
“the fibrous tissue not only being along the vessels between the
lobules, but extending in between the individual cells, the organ
being enlarged and having a smooth or more rarely a finely granu-
lar surface.” In the animals examined by him in Canada there

2 Report of the Director of Agriculture of the Cape of Good Hope, 1906, p. 41

Montreal Medical Journal, February, 1902.
w,w.-,.+, +‘ + +‘ , , , .,,‘.

ACTION OF SENECIO ALKALOIDS 533

was abundant production of thin bile and the gall-bladder was

generally very full and the faces well stained. The abdominal
lymph glands were generally large and succulent and moderate
ascites was present, with gelatinous cedema of the mesenteries
and walls of the intestines. Numerous follicular ulcers were
found in the fourth or true stomach but were generally in a cica-
trized condition except in very acute cases. In animals killed
in early stages, the most noticeable features are fatty degenera-
tion of the liver cells with great congestion of the liver vessels
(Johnston). This stage appears to be succeeded by rapid de-
struction of the liver cells and replacement of these by delicate
new connective tissue. Johnston and Adami isolated an organ-
ism from the organs, but were unable to induce cirrhosis in ani-
mals inoculated by it and it subsequently proved to be the colon
bacillus. Johnston and Hammond showed that the blood of
cattle affected with the disease agglutinated the organism isolated
from the liver, which apparently confirmed their view that this
organism is the cause of the disease. The cogency of this argu-
ment has been lessened however by the observations of Ford4
that in over 80 per cent of the livers of normal animals bacteria
may be found. Adami (1902) still appears to hold that there is a
direct relationship between these organisms and the cirrhotic
process, but considers that they are not the primary cause of the
disease, but that, gaining entrance through a gastric ulcer or
other inflammatory condition they induce the characteristic
changes in the liver.
There has always been a popular impression in Pictou that
the disease was associated with the ragwort, but the earlier
experiments (1882) seemed to negative this idea. In 1902, Gil-
ruth5 was led to examine the effects of the Senecio jacobcea of New
Zealand by the observation that cattle on a station affected by
Winton disease had been eating this plant. He fed two healthy
calves with a rafion containing this weed, and found that they

Journ. of Hygiene, i, p. 277, 1901.

‘Tenth Report of the Department of Agriculture, Wellington, New Zealand,
p. 300.
534 ARTHUR R. CUSHNY

became ill after about eighteen days and died on the twenty-
eighth and thirtieth day of the experiment. The symptoms
were typical of Winton disease. Post mortem the peritoneal
cavity contained straw colored fluid and the gall-bladder was
full of dark fluid of the consistency of treacle. The connective
tissue of the liver was much increased especially around the
portal canal and smaller interlobular branches of the portal
vein; from these bands of finer fibres extended between the liver
cells sometimes isolating them completely. Areas of congested
portal capillaries were found irregularly in the sections. The
liver cells were distorted in shape and atrophied. The capsule
of the liver was thickened and trabecula passed from it into the
tissue.
In South Africa, Chase6 found that the inoculation of the bile,
blood and stomach contents of an animal dead of Molteno dis-
ease had no effect upon healthy cattle. After Gilruth’s results
in New Zealand appeared, the &necio burchellii of South Africa
fell under suspicion and Chase found that about 4 oz. of this
plant given on four successive days to an ox caused his death with
the usual symptoms and the typical post mortem appearances on
the fifth day. He does not state whether the plant was given
green or dry.
Dixon notes (1906) that the Senecio latifolius grows abundantly
in the area in which the hepatic cirrhosis is observed in the East
London district, South Africa.
In Canada Pethick7 carried out an admirably planned investi-
gation of the subject in 1903; sixteen head of cattle were kept in
a new stable and fed on hay from Pictou, while sixteen controls,
kept in an old stable in which several cattle had previously died
of the disease, received hay from Quebec, where the disease is
unknown. These sixteen controls remained healthy for twenty-
three months, though they were placed in contact with the dis-
eased animals and in some cases were inoculated with the blood
or ascitic fluid of animals that had died of the disease; when

‘bc. cit.
Department of Agriculture, Canada. Health of Animals: Special Report on
Pictou Cattle Disease, Ottawa, 1906.
 ACTION OF SENECIO ALKALOIDS 535

slaughtered at the end of twenty-three months their organs were

free from disease. Of the sixteen cattle to which hay from Pic-
tou was supplied, fifteen died of the disease as was verified by the
post mortem examination, and the sixteenth animal was found to
present the characteristic lesions of the disease when it was
slaughtered at the end of twenty-six months. One animal fed
with chopped ragwort mixed in Quebec hay died in twelve months
of acute cirrhosis, while the control fed on Quebec hay remained
healthy. It was found that sheep can be fed on ragwort with
comparative impunity, though they also suffer after some time
and the flesh finally assumes a yellowish tint (jaundice?). Sheep
may thus be employed to eat down the ragwort on farms infected
with the weed and thus to extirpate it. Rutherford8 states that
sheep eat ragwort with impunity whether the plant is in the green
or in the dry state. Cattle refuse to eat it in the green st+ate and
the poisoning seen in Canada arises only from the dried ragwort
in the hay. But cattle in South Africa and New Zealand eat
the green plant also when other fodder is scarce. (Chase.)
The experiments of Gilruth, Chase and Pethick show beyond
doubt that poisoning with different species of Senecio is the cause
of the disease known as Pictou disease in Canada, Molteno disease
in South Africa and Winton disease in New Zealand. The spe-
cies of Senecio hitherto incriminated are S. jacobcra in New Zealand
and Canada, and S. burchellii and probably S. latifolius in South
Africa, but it is possible that other species may be equally poison-
ous. Curiously enough the S. jacobaa which grows in abundance
all over the eastern districts of England and Scotland (from which
it was introduced into Canada) is not known to cause any symp-
toms in cattle, and the S. vulgaris or groundsel which is found
equally widely distributed is also reputed to be harmless.
As regards the view that the hepatic cirrhosis is a bacterial dis-
ease, the experiments given above indicate that the role played by
the microbes is a secondary one, which bears the same relation
to the disease as the terminal infection occurring in other chronic
wasting diseases.
8 Report of the Veterinary Director General and Live Stock Commissioner,
Ottawa, 1909, p. 21.
536 ARTHUR R. CUSHNY

The chemistry of the Senecio genus has not received much

attention. Grandval and Sejour9 found in the common ground-
sel (S. vulgaris) two alkaloids, senecionine (C18H2506N) and sen-
ecine, but make no statement as to their toxicity. Dr. H. E.
Watt1#{176}has recently examined the S. latifolius of Cape Colony,
which is believed to be responsible for at any rate some of the out-
breaks of disease in South Africa, and has succeeded in isolating
from it two alkaloids, which in the crude state amounted to 1.72 per
cent of the dried plant gathered before flowering and 0.76 per
cent in those gathered after flowering. One of these senecifoline
corresponded to the formula C18.H2708N, forms crystalline salts
of which the nitrate melts and decomposes at 240, the hydro-
chloride at 260, and the aurichloride at 220; it can be decomposed
into an acid and base by alkali and both of these products were
isolated and examined. The second alkaloid, senecifolidine,
C18H,507N, also forms crystalline salts of which the nitrate melts
at 145.
Professor Wyndham R. Dunstan, under whose direction this
chemical investigation was carried out kindly sent me the nitrates
of these bases for pharmacological examination.

SENECIFOLINE NITRATE

Injected into frogs in quantities of 10 to 40 mgs. senecifoline

induced no symptoms whatever for five to fifteen days, and the
quantity given seemed to exercise no influence on the interval
between the injection and the onset of symptoms. At the end
of this time the condition varied in different animals. In some
the position was fairly normal, but the respiration had ceased
and gaping movements were made apparently in attempts to
vomit, for sometimes the stomach was protruded through the
mouth. In others clear blood escaped from the mouth and the
animal showed all the symptoms of collapse from hamorrhage.
In others strychnine-like spasms and muscular twitching were

‘Compt. rend. de l’aead. des Sciences, Paris, 120, p. 1120, 1895

10 Trans. Chem. Soc., 95, p. 466, 1909.
. + ,. - + ++

ACTION OF SENECIO ALKALOIDS 537

present. Death followed in a few hours after the first symptoms

developed. The appearances post mortem varied greatly in
different frogs. In some the abdominal cavity was found to
contain quantities of clear fluid; the stomach was often found to
be normal or somewhat congested, but hamorrhage had occurred
in the bowel which sometimes contained much blood-stained
mucus; the cloaca and lower bowel often were filled with a mixture
of blood and mucus. Hamorrhage from the lungs was the chief
abnormal feature in several frogs.
A number of experiments were performed on cats, white rats
and rabbits, and the reaction of these animals to the alkaloid was
uniform in most particulars. The drug was generally injected
hypodermically but the same results were obtained when it was
given by means of the stomach tube in cats.
In the cat senecifoline nitrate induces two sets of symptoms
which are quite distinct in character.
Acute symptoms. After a small dose, e.g., 0.01 to 0.02 gm.
per kg. has been injected hypodermically there is often some
salivation which continues for half an hour or longer. After
large quantities, e.g., 50 mgs. per kg. this feature is more marked,
the saliva pouring from the mouth and later falling in long strings
from the jaws. The pupils may be somewhat dilated, but these
symptoms disappear in the course of twenty-four hours and apart
from one or two stools of rather loose consistency the animal
appears to have recovered completely until it shows the symp-
toms described later. In. somewhat larger doses vomiting is
often induced in the course of the first two hours. When quan-
tities of 0.1 to 0.2 gm. per kg. are injected into cats, the profuse
salivation is accompanied by vomiting, the pupils are widely
dilated, the respiration is extremely accelerated (up to 300 or.
more per minute). The animal shows the sudden shrinking
movements, which are often seen under cocaine and which have
been construed as indicating hallucinations, and somewhat later
violent clonic convulsions may be developed. These resemble
closely those developed under cocaine and other convulsant poi-
sons, periods of very active movement alternating with pauses
during which the animal lay still and in which the respiration was
538 ARTHUR R. CUSHNY

slow at first but gradually accelerated until it culminated in a

renewed seizure. All of these symptoms pass off in the course
of two hours or less after the injection and next day the animal
appears fairly normal; the appetite may seem poor or some depres-
sion may be present but very often no further symptoms can be
made out for several days.
In the rabbit and rat I have not been able to elicit any pri-
mary symptoms of so marked a character. In the rat even 1.0
gm. per kg. caused no convulsions but there did seem to be some
excitement, the animal running about more, climbing up its cage
and making gnawing movementh. The respiration was also
accelerated at first, but it was difficult to determine how far this
was due to the injection as such and how far to the specific action
of the drug. Smaller doses of senecifoline nitrate were followed
by no symptoms whatever for many hours after the injection
and then gave rise to the characteristic changes which will be
described below.
These acute symptoms point to an action of senecifoline on the
central nervous system similar to those observed under a number
of other convulsant poisons. The salivation in cats is the most
common feature and appears to be merely an accompaniment of
the nausea which under larger doses causes vomiting. The accel-
eration of the respiration resembles that seen under cocaine and
the convulsions and apparent hallucinations are also commonly
elicited by this drug and its congeners. The acute symptoms are
thus referable to stimulation of the upper part of the central ner-
vous axis similar to that induced by cocaine and apomorphine.
No distinct reflex in the spinal reflex irritability could be made out,
so that the action is confined to the medulla oblongata and higher
centres.
In one cat anaesthetized with paraldehyde a blood pressure
experiment was performed and the salivary secretion was deter-
mined at the same time by a cannula inserted in the submaxil-
lary duct.. The intravenous injection of small quantities 0.01 to
0.025 gm. had little or no effect on the blood pressure or rate of the
heart. Larger amounts, 0.05 to 0.15 gm. induced a rapid fall in
blood pressure followed by a return to the normal, which was

a
 + . .;

ACTION OF SENECIO ALKALOIDS 539

reached in two to five minutes. The salivary secretion was unaf-

fected throughout and the pupil remained unchanged in size.
The local application to the eye of the cat of senecifoline nitrate
in 5 per cent solution caused neither irritation nor anaesthesia and
had no effect on the size of the pupil. The dilatation of the
pupil observed among the acute symptoms is therefore to be
ascribed to the general excitement as in the case of many other
excitants.
The acute symptoms seem to arise from a stimulant action on
the upper part of the centfal nervous axis. As in the case of
many other poisons acting in this way the symptoms are much
more readily elicited in the cat than in the rodents.
Subacute Symptoms. After the recovery from the primary
symptoms cats remained to all appearance quite normal for at
least twenty-four hours and often for three days up to a week.
After small doses, e.g., 0.07 gm. per kg., salivation was generally
absent and no symptoms developed at all for several days. In
those remaining without symptoms for the longer periods, some
loss of weight was generally noted, but this was often insignificant
until the more severe subacute symptoms set in. These were
often introduced by a stool of rather loose consistency, and by
loss of appetite, the animal eating nothing or often vomiting
what little it had swallowed. Weakness and disinclination to
move soon appeared, and somewhat later when the animal was
liberated from its cage and encouraged to walk, the gait was
unsteady and staggering; the legs were sometimes curiously stiff
and wide apart. When sitting still the body swayed to and fro,
the head fell lower than normally and though it was raised when
the animal was disturbed it assumed the same position very soon.
This condition of apathy soon deepened into stupor in which the
animal lay still and could be aroused with difficulty or somewhat
later failed to respond to sounds or touch. The respiraton
became gradually slower, the temperature fell 4 to 5 below nor-
mal, the pulse was also slow. Finally the respiration gradually
ceased, the heart continuing to beat for some time afterwards
and few or no convulsive movements occurring at death. In
one experiment on the cat a large stool consisting of almost
540 ARTHUR R. CUSHNY

unaltered blood was passed before death and at the autopsy

blood was found to have exuded from the rectal mucous mem-
brane in large quantity. The urine examined during these symp-
toms sometimes contained a trace of protein, but was generally
free from it.
These symptoms succeeded each other rapidly, death occurring
about twenty-four to forty-eight hours after their first appearance
in the majority of cases. Thus an animal often remained appar-
ently normal for four to five days after the drug was administered
but died on the day after symptoms first set in. In some experi-
ments in which a large dose was injected, death occurred in two
days, and in these the weight was found to have fallen about 10
per cent, while in animals surviving four to five days, as much as
20 per cent of the original weight had been lost. Even in animals
which recovered after senecifoline injection some loss of weight
was often noted.
The smallest dose which proved fatal to the cat was 0.016 gm.
per kg. As a general rule the dose used in the cat was about
50 mgs. per kg., and this amount proved fatal in four to six days
whether it was injected hypodermically or given by the stomach . +

tube.
In the rat the symptoms were very similar, setting in with
loss of appetite (but with no diarrhoeic stool), after which increas-
ing weakness and deepening stupor and coma followed; the res-
piration became slow and finally ceased, the heart beating for
some minutes longer. The senecifoline nitrate was always in-
jected hypodermically, and great variation was found in the
quantity which proved fatal in different animals. Thus in one
rat 0.044 gm. per kg. proved fatal in three days, while others sur-
vived 0.135, 0.16 and even 0.22 gm. per kg. The interval between
the injection and death again showed great variation, some
rats dying within twenty-four hours of the injection, while others
survived five to six days. The amount of alkaloid injected did
not seem to bear any definite relation to the time of survival. In
rabbits the symptoms resembled those in cats, and the fatal
dose was not ascertained in them.
 ACTION OF SENECIO ALKALOIDS 541

The post mortem appearances differed considerably in differ-

ent animals of the same species even when they had received the
same amount of the poison and had survived the same time. In
the cat there was generally an unusual amount of fluid in the
abdominal cavity and this was sometimes of a bright yellow
color. In one cat in which there was marked jaundice the
abdominal cavity was filled with fluid of a deep bile color. Some
congestion of the great omentum was often present and this was
sometimes very marked; small ecchymoses were present in the
omentum and fat deposits in one or two animals. The stomach
generally contained black masses of half digested blood espe-
cially towards the pyloric orifice, and the duodenum also contained
some effused blood mixed with mucus, or sometimes was filled
with black masses of blood; congestion of the stomach and upper
part of the intestine was generally to be made out, but sometimes
there was no abnormal appearance in the stomach or bowel
except some slight congestion of the vessels.
The liver was swollen and congested and bled freely when cut
into. A section presented areas of dark and light color. The
gall bladder was distended with very dark colored viscous bile
which could only be expressed from it with difficulty. The
kidneys sometimes presented the appearance of fatty degenera-
tion. Small hamorrhages were often found in various organs,
such as the omentum, lungs, pancreas, and were present in most
cases in the stomach and intestine. In the rabbit, similar lesions
were found at the autopsy. In one case a quantity of brownish
fluid was found in the pleural and pericardial cavities as well as in
the abdomen.
In the rats .hamorrhages into the stomach and duodenum were
almost invariably found and the contents of the upper part of the
gut were thick mucus, stained pink with blood. The gastric
hamorrhage was confined to the pyloric half of the organ or
true stomach. In the peritoneal cavity, bile stained or blood
stained fluid was often present, sometimes in large amount, and
in several instances fluid was present also in the pleural and pen-
cardial cavities. The tissues were rarely jaundiced. The liver
was generally enlarged and soft and contained large areas of
542 ARTHUR R. CUSHNY

intense congestion. Hamorrhages occurred in a number of

organs. The kidneys generally appeared normal. In the rat
as in the cat, the thoracic organs were generally normal in appear-
ance, save for occasional discrete hamornhages in the lungs and
rare pleural effusions.
Dr. C. Bolton kindly examined some of the organs microscopi-
cally for me, and found marked congestion and hamonrhages in
the liver; the hamornhage in some cases was confined to the peri-
pheral half of the lobules, which were mapped out by it, in other
instances both hepatic and portal veins seemed to be equally
involved. When the peripheral part of the lobule alone was
affected, the hepatic cells in the centre were normal in appear-
ance, but further outwards they became distorted by the blood
cells and stained badly and towards the interlobular vein they
were quite colorless and evidently in process of disintegration.
Large areas containing many lobules were necrosed, the liver cells
not taking the stain at all, and here hiemorrhages were also pres-
ent though not in greater degree than in those parts in which the
hepatic cells were in part preserved, so that this necrotic action
appeared to be independent of the hlemorrhages. In acute
poisoning the liver cells often contained globules of fat, but this
did not seem to be in excess of that seen in normal animals. In
less acute cases this fat was not present. There was considerable
infiltration of round cells round the portal canal, especially involv-
ing the smaller bile ducts and often occupying their lumen and
also extending outwards from the ducts between the liver cells;
and this was present in subacute cases though it was more marked
in chronic poisoning. Catarrh of the smaller bile ducts was indi-
cated by shedding of epithelium.
The liver symptoms suggested that senecifoline might have
some of the properties of toluylendiamine, and its effect on the
blood was therefore examined. I may mention that the blood
clotted normally in animals poisoned with the alkaloid, except
when marked jaundice was present and the peritoneal fluid also
clotted after standing sone time. There was no hsemolysis in
the blood drawn from the heart before or after death in animals
poisoned with senecifoline, and no hamog1obin appeared in the
 ACTION OF SENECLO ALKALOIDS 543

urine. Normal cat’s blood diluted with 9 parts of saline solution

was put in the incubator at 38 with a varying proportion of seneci-
foline nitrate. After sixteen hours a very slight degree of lysis
had occurred in the tubes containing senecifoline, while the con-
trols remained unchanged. In the senecifoline tubes there was
also some methamoglobin and this later passed into dark amor-
phous insoluble masses. This lysis and methamoglobin forma-
tion was fairly obvious in tubes which contained 1 of senecifoline
nitrate in 300 and was barely visible when the concentration was
1:1400. Strychnine nitrate added to control tubes and exposed
in the same way had no effect on the blood pigment and produced
no lysis. So that this action of senecifoline cannot be attributed
either to the nitrate half of the molecule or to the weakness of the
base inducing an excess of H ions. Senecifoline thus has a weak
hamolytic action and leads to methamoglobin formation in the
shed blood, thus resembling some of the oxidizing inorganic salts.
The concentration necessary to elicit this is much higher than
that reached in the living tissues and in the intact animal the blood
was not found to be lysed nor to contain methalmoglobin, so that
the subject was not pursued further. It may be mentioned how-
ever that some of the oxidizing poisons, such as the iodates,
induce similar lesions in the rat’s stomach to those described
under senecifoline.
Chronic poisoning was induced in two experiments in order to
compare the symptoms elicited with those observed in cattle in
Pictou disease.
A young cat of 1200 gms. weight.

5. II. 30 mgs. senecifoline nitrate hypodermically.

9. II. Slight diarrhoea.
21. II. Normal. 20 mgs.
22. II. Normal. 20 mgs.
23. II. Normal. 25 mgs.
24. II. Normal. 25 mgs.
25. II. Normal. 25 mgs.
7. III. Normal. 75 mgs.
10. III. Normal 75 mgs. (No more drug given after this.)
Weight, 980 gms., salivation, sits with closed eyes, ate
very little and vomited once.
544 ARTHUR R. CUSHNY

11. III. Ate some meat, but sits still, weak and apathetic; saliva-
tion, tears and snuffles.
14. III. Eats fairly well, but salivates a good deal in doing so.
No jaundice. Some conjunctivitis. Weight, 875 gms.
16. III. Did not eat yesterday or today. Very thin but fairly
active. Weight, 820 gms.
17. III. Did not eat meat but drank some milk. Weight, 790
gms.
18. III. Found dead in the morning. Weight, 770 gms.

An examination was made at once. No subcutaneous fat was

present, no jaundice and no marked congestion. One to two
cc. of clear fluid in the peritoneal cavitSr. The pyloric end of the
stomach contained a quantity of black clotted material, which was
not attached to the mucous membrane. The duodenum had
some clear glairy fluid, while the ileum contained black masses
and the large bowel its usual contents. No erosions could be
made out anywhere along the tract to account for the blood in
the stomach. The gall-bladder was distended with a very light
colored fluid with white masses in it. Liver weighed 35 gm.;
spleen 34 gm., and kidneys together 14 gm. Lungs and heart
seemed normal and the blood in the heart presented no abnormali-
ties and was not lysed. HiStological examination of the liver
showed that organ to be in an advanced state of degeneration.
The liver cells had disappeared in great part and the few to be
seen stained poorly. The great bulk of the picture was occupied
by blood corpuscles and their debris. Round the vessels were
masses of round cells which appeared to be in process of change to
connective tissue. This round cell inifitration extended also into the
remains of the lobules and between the surviving liver cells. The
cirrhosis had not proceeded as far as is described in cattle, but
was of the same nature, and on the other hand was obviously
a further development of the process seen in cats which had died
from a single dose of senecifoline. In the kidney a number of
small halmorrhages were found and many of the cells of the con-
voluted tubules were swollen and stained badly.
A rat of 255 gms. received subcutaneously 5 mg. of seneci-
foline nitrate on the 1st, 9th, 11th, 18th, 21st (weight, 217 gms.)
+ . . + + . + +. ++

ACTION OF SENECIO ALKALOIDS 545

26th (weight, 205 gms.), 29th (205 gms.), 32d (209 gms.), 35th
(195 gms.), and 40th day of the experiment and was then found to
weigh 168 gins., having lost about a third of its original weight.
The animal was now very weak, with labored respiration, and ate
very little. The coat was very rough, the hairs standing up
erect and separate. It was found dead on the morning of the
41st day of the experiment. The abdominal vessels were found
congested, and in the upper part of the small intestine small
haemorrhages, some of which were quite recent, were seen. The
stomach seemed normal and had no halmorrhages. The liver
weighed 8 gms., did not appear congested, and macroscopically
nothing abnormal was seen in it. Some hepatization and several
small abscesses were found in the lungs. Microscopically the
cells of the kidney tubules were found so swollen as to close the
lumen entirely. The liver showed large areas of necrosis and,
between these, groups of lobules in which the liver cells stained
badly, while the connective tissue was increased in amount and deli-
cate strands extended between the individual cells. Some ham-
orrbages had occurred in the liver and the whole organ was much
congested.
SENECIFOLIDINE NITRATE

Only a small quantity of this alkaloid was available, and this

was used for experiments on rats. Its effects resembled those of
senecifoline, the same changes occurring in the liver and alimen-
tary tract. As far as could be ascertained with the small quan-
tity at my disposal, the two alkaloids seemed to be equally toxic
in the rat.

DISCUSSION AND SUMMARY

The symptoms and post mortem findings in animals poisoned

with these alkaloids resemble so closely those described by Gil-
ruth, Chase, Pethick and others, in cattle and horses, that there
can be no question that the cause is the same in each and that the
Pictou, Winton or Molteno disease is really more or less chronic
poisoning with the Senecio alkaloids. It is true that the hepatic
cirrhosis in my experiments was not so complete as seems to have
546 ARTHUR R. CUSHNY

been observed in cattle, but this is sufficiently explained by the

shorter duration of the intoxication in the smaller laboratory
animals; and the initial stages of cirrhosis were certainly present
in the two instances of chronic poisoning which I have given.
The dominating lesion present in acute and subacute poison-
ing is hamorrhage which may occur in almost any organ but is
constant in the liver and almost invariably present in the stom-
ach and bowels. This is accompanied by fatty changes and
necrosis of the liver cells and by great congestion of the organ. I
am unable to determine which is the primary feature, whether the
cell changes induce the congestion and hamorrhage or vice versa.
Very often areas of congestion alternated with areas in which
the hepatic cells were swollen and stained badly or contained
large globules of fat, and it appeared as if the two processes of
congestion and parenchymatous change were independent of
each other. These liver changes account for many of the symp-
toms and post mortem changes noted. Thus the thick viscous
bile found in the gall-bladder may well be secondary to the hannor-
rhages. A similar bile is secreted in poisoning with toluylen-
diamine and this has been shown by Stadelmann” to arise from
the excessive bile pigment formation from the hamolysis occur-
ring in the liver. In senecifoline poisoning the large hamor-
rhages in the liver may have the same result as the hamolysis in
toluylendiamine poisoning, the bile pigment beingincreased and the
bile becoming correspondingly dark and viscous. In some of our
cases this gave rise to jaundice of greater or less intensity and
this symptom has been observed in cattle poisoned with ragwort;
the yellowish color of sheep muscles after feeding on ragwort
appears to be another example of this action.
The dropsy also arises in all probability from the hepatic
hamorrhages, obstructing the flow from the portal system.
This was very marked in some of my experiments, and in many
of them the peritoneal fluid was increased. Dropsy has also
been observed in cattle poisoning.

11 Joannovics and Pick: Zeitsch. f. exp. Path. u. Ther., vii, p. 185.

 ACTION OF SENECIO ALKALOIDS 547

The loss of appetite and late vomiting in my experiments are

sufficiently explained by the hamorrhages in the stomach and
by the general disturbance of nutrition from the hepatic changes.
Bleeding from the intestine was almost constant, in almost all
cases in the duodenum, in one from the rectum. Cattle appear
to pass more blood by the bowel than the laboratory animals
but this appears to be merely a question of degree.
The experiments hitherto detailed were performed with the
alkaloids of Senecio latifolius, which, as has been said, is held
responsible for some of the epidemics in South Africa, and my
results indicate that these alkaloids are capable of inducing the
symptoms and lesions characteristic of the disease. The Senecio
jacobcva which has been shown to be responsible for the disease in
New Zealand and Canada, grows in profusion in England and Scot-
land, but enquiries made in various parts of the country indicate
that poisoning with this plant and hepatic cirrhosis are unknown
here. One hundred grammes of dried Senecio jacoba3a (ragwort)
collected in England for me in June and August were extracted
with alcohol, the alcoholic solution evaporated at a low tempera-
ture and the residue taken up with water acidulated with hydro-
chloric acid. This was neutralized and injected hypodermically
into a cat, but induced no symptoms whatever. The extract
prepared in the same way from 200 gms. dried ragwort similarly
gave no results when given by the stomach tube to a cat. Senecio
jacobxa grown in Canada and dried was extracted in he same
way by Professor Dunstan and sent to me for examination. A
quantity corresponding to 20 gms. of the dried herb was injected
into a cat hypodermically without result. If the alkaloid con-
tent of the dried herb were even as low as one per cent this dose
would have contained 0.2 gin, which is certainly poisonous.
These results would therefore seem to indicate that the S. jacobaa
is devoid of the toxic properties of S. latifolius, whether the
plant s grown In England or in Canada. This is however incom-
patible with the results of Gilruth and Pethick, who showed
definitely that the disease in Canada and New Zealand is due
to this species. The discrepancy between these results and mine
may probably arise either from the plant from which my prepara-
548 ARTHUR R. CUSHNY

tions were made having been collected at the wrong season, or

possibly from the poisonous principle having undergone change
into some inert form in the course of preparation or drying.
Senecio sylvaticus collected in Yorkshire in August proved
equally inactive. Senecio vulgaris or common groundsel collected
in England and prepared in the same way proved poisonous:
a cat which received the extract from 2 gms. of the dried plant
dying in ten days with symptoms resembling those arising from
senecifoline, but with more marked diarrhoea. The post mortem
findings were similar to those observed from senecifoline.
I hope to investigate further the toxicity of S. jacoba?a with the
hope of elucidating the curious discrepancies between my results
and those of Gilruth and Pethick.