%\pdfoutput=1\relax\pdfpagewidth=8.26in\pdfpageheight=11.69in\pdfcompresslevel=9

\input amstex
\documentstyle{amsppt}
\loadmsbm
\hcorrection{18mm}  \vcorrection{5mm}
\nologo \TagsOnRight \NoBlackBoxes
\headline={\ifnum\pageno=1 \hfill\else%
{\tenrm\ifodd\pageno\rightheadline \else
\leftheadline\fi}\fi}

\def\rightheadline{EJDE--2001/13\hfil Multiple positive solutions  \hfil\folio}
\def\leftheadline{\folio\hfil G. L. Karakostas \& P. Ch. Tsamatos
 \hfil EJDE--2001/13}

\def\pretitle{\vbox{\eightrm\noindent\baselineskip 9pt %
 Electronic Journal of Differential Equations,
Vol. {\eightbf 2001}(2001), No. 13, pp. 1--10.\hfil\break
ISSN: 1072-6691. URL: http://ejde.math.swt.edu or http://ejde.math.unt.edu
\hfill\break
ftp ejde.math.swt.edu (login: ftp)\bigskip} }

\topmatter
\title
Multiple positive solutions for a nonlocal boundary-value problem with
response function quiet at zero 
\endtitle

\thanks
{\it Mathematics Subject Classifications:} 34B18. \hfil\break\indent
{\it Key words:} Multiple positive solutions, nonlocal boundary
value problems, \hfil\break\indent
functions quiet at zero, Krasnoselskii's fixed point theorem.
\hfil\break\indent
\copyright 2001 Southwest Texas State University. \hfil\break\indent
Submitted January 23, 2001. Published February 19, 2001.
\endthanks

\author  G. L. Karakostas \& P. Ch. Tsamatos  \endauthor

\address G. L. Karakostas  \hfill\break\indent
Department of Mathematics, University of Ioannina,
451 10 Ioannina, Greece
\endaddress
\email gkarako\@cc.uoi.gr
\endemail

\address P. Ch. Tsamatos    \hfill\break\indent
Department of Mathematics, University of Ioannina,
451 10 Ioannina, Greece
\endaddress
\email  ptsamato\@cc.uoi.gr \endemail

\abstract
The existence of positive solutions of a nonlocal boundary value problem for a
second order differential equation is investigated.  By assuming that the
response function is quiet at zero, in a sense introduced here, and it satisfies
some easy conditions, existence results for a countable set of positive
solutions are given.  
\endabstract \endtopmatter

\document

\head 1. Introduction \endhead

In a recent paper the authors gave sufficient conditions for the existence
of a positive solution of
the nonlocal boundary value problem
$$ \gather
(p(t)x')'+q(t)f(x)=0,\enskip \hbox{a.a.}\enskip t\in [0,1]\tag 1.1 \\
 x(0)=0,\tag 1.2 \\
x'(1)=\int_{\eta}^{1}x'(s)dg(s)\tag 1.3 
\endgather
$$
where $\eta \in (0,1),$ see [25]. Among these conditions, the monotonicity
of the response
function $f$ seemed to be crucial in the proof. In this paper we 
weaken the
monotonicity condition on $f$ by assuming that this function is
quiet at zero in the
following sense: Given any pair of sequences $(x_n )$, $(y_n )$ with $0\le
x_n\le y_n$
converging to zero it holds $f(x_n )=O (f(y_n ))$, (where $O$ stands for
the big-$O$
symbol). This definition, which is introduced here, refers to functions $f$
which do not
vanish at least on $(0, +\infty )$. It is not difficult to see that if
$f(0)>0$, or, if $f$ is
increasing in a right neighborhood of zero, then $f$ is quiet at zero.
\par
Moreover we extend the results of [25] and show that our boundary value
problem can
admit a countable family of positive solutions.

Here we have to mention that boundary value problems of the form
$(1.1)$, $(1.2)$,
$(1.3)$ are mainly motivated by the works of Bitsadze [8], Bitsadze and
Samarskii [9]
and Il'in and Moiseev [23] and includes as special cases multipoint
boundary value problems
considered in [19] and [20]. Moreover, the authors in [25-28] proved
recently existence
results for some relative nonlocal boundary value problems. On the other
hand the problem
of the existence of multiple solutions (at least two) for various types of
boundary value
problems is recently the subject of many papers. Among others we refer to
[1-4, 6, 7, 10-12, 14-17, 21, 22, 24, 29, 32]. The technique in these
papers is based on
fixed point results in cones. Most of them are based on the following well
known
fixed point theorem due to Krasnoselskii [30].

\proclaim{Theorem 1.1}
Let ${\Cal B}$ a Banach space and let  $\Bbb K$  be a cone in ${\Cal B}$.
Assume $\Omega _1 $,
$\Omega _2 $ are open subsets of $E$, with $0\in\Omega _1 \subset \overline
{\Omega _1
}\subset \Omega _2$, and let
$$A\colon  \Bbb K\cap (\Omega _2\setminus \overline {\Omega _1 } )\to \Bbb K$$
be a completely continuous operator such that either
$$\|Au\|\le \|u\|,\enskip u\in \Bbb K \cap \partial \Omega _1 ,\enskip
\|Au\|\ge \|u\|,\enskip
u\in \Bbb K \cap \partial \Omega _2 $$
or
$$\|Au\|\ge \|u\|,\enskip u\in \Bbb K \cap \partial \Omega _1 ,\enskip
\|Au\|\le \|u\|,\enskip
u\in \Bbb K \cap \partial \Omega _2 .$$
Then $A$ has a fixed point in $\Bbb K\cap (\Omega _2\setminus \overline
{\Omega _1 } ).$
\endproclaim
In this norm form and in its degree form (see [18]), Theorem 1.1 is applied in
[1, 2, 3, 11, 14-17, 22, 24, 29, 32]. Some interesting versions of the
theorem (see, e.g.,
[5, 13, 31]) are also applied in [4, 6, 7, 12, 21]. Finally, we mention
that by using a different
fixed point theorem due to Ricceri [33] a multiplicity existence result is
obtained in [10].
\par Here we apply Theorem 1.1 to obtain existence results for a countable
set of positive
solutions of the boundary value problem $(1.1)$, $(1.2)$, $(1.3)$, where
the main hypothesis is
that the function $f$ is a quiet at zero function. This meaning is given in
the following
section.


\head 2. Quietness at zero\endhead

We introduce the following definition:
\par {\bf Definition.} A continuous function
$f:[0,+\infty )\to\Bbb R$, with $f(x)>0$ when $x>0$, is said to be quiet at
zero, if for any pair of
sequences $(x_n )$, $(y_n )$  with $0\le x_n\le y_n$, $n=1,2,...$, which
converge to zero, it
holds $$f(x_n )=O (f(y_n )).$$
\par
This means that there is a $K>0$ such that $f(x_n )\le Kf(y_n )$ for all $n$.
An equivalent form of this definition, which will be used in our proofs, is
given by the
following lemma:
\vskip 0.2cm
\proclaim{ Lemma 2.1} A continuous function  $f:[0,+\infty )\to \Bbb R$,
with $f(x)>0$
when $x>0$ is quiet at zero, if and only if for each $T>0$ there is a
$\mu\ge1$ such that
for all $\tau\in(0,T)$ it holds
$$\sup\{f(x):x\in [0,\tau ]\}\le\mu\inf\{f(x):x\in [\tau ,T]\}.\tag 2.1$$
\endproclaim
\demo{ Proof}
Assume that $f$ is quiet at zero and there is a $T>0$ such that for each
positive integer $\mu$
there is a point $\tau_{\mu}\in (0,T)$, with
$$\sup\{f(x):x\in[0,\tau]\}>\mu\inf\{f(x):x\in [\tau ,T]\}.$$
This implies that there are sequences $(x_{\mu})$, $(y_{\mu})$, with
$0<x_{\mu}\le \tau_{\mu}\le y_{\mu}$ and $f(x_{\mu})>\mu f(y_{\mu})$ for
all $\mu$.
Since $f$ is bounded on $[0, T]$, taking limits as $\mu\to +\infty$, we get
$f(y_{\mu})\to
0$, thus $y_{\mu}\to 0$ and, so, also $x_{\mu}\to 0$. These facts
contradict to our
assumption.
\par
For the "if" part of the proof, we assume that there are sequences $(x_n
)$, $(y_n )$, with
$0\le x_n < y_n$ and $x_n\to 0$, $y_n\to 0$ and, moreover,
$$\displaystyle{\lim_{n\rightarrow\infty}}\frac{f(x_n )}{f(y_n )}=+\infty
.\tag 2.2$$
Set $T:= max\{y_n\}$. Then, by assumption, there is a $\mu\ge 1$ such that
for all $\tau\in
(0,T)$ it holds $(2.1)$. From $(2.2)$ there is an index $n_0$ such that
$f(x_{n_ {_0}})\ge
(\mu +1)f(y_{n_{_0}}).$ Set $\tau:=x_{n_{_0}}$ and observe that
$$ \aligned \sup\{f(x): x\in [0,\tau ]\}&\ge f(x_{n_{_0}})\ge (\mu
+1)f(y_{n_{_0}})\\
&\ge (\mu+1)\inf\{f(x):x\in [\tau ,T]\},
\endaligned
$$
contradicting to 2.1. The proof is complete.
\qed\enddemo
{\bf Remark.} We observe that if $f(0)>0$, then $f$ is quiet at zero.
Indeed, for each
$T>0$ the real number
$$\mu:= \frac{\sup\{f(x):x\in [0,T]\}}{\inf\{f(x):x\in [0,T]\}}$$
works in $(2.1)$ for all $\tau\in (0,T)$.
\par
Also, if $f$ is a nondecreasing in a right neihborhood of zero, then it is
quiet at zero. To see this
we assume that $f$ is nondecreasing on an interval $[0,\delta ]$ and
consider two sequences
$(x_{n}), (y_{n}),$ with $0\le x_n < y_n$ for all $n$ and $x_n\to 0$,
$y_n\to 0.$ These sequences
belong eventually in the interval $(0,\delta ],$ hence, by the monotonicity
of $f$,
$f(x_{n})\le f(y_{n})$ for all large $n$. This proves that the function $f$
is quiet at zero.
\par
\vskip 0.1cm
{\bf Example.} We give a simple example of a function, which is not quiet
at zero. Consider
sequences $(x_n )$, $(y_n )$ such that $x_n\to 0$, $y_n\to 0$, with $0<x_n
<y_n <x_{n-1}
<y_{n-1}<...<x_1 <y_1$ and $\lim\frac{f(x_n )}{f(y_n )}=+\infty.$ Then,
define a new sequence
$(z_n)$ with $z_{2n}=x_n$ and $z_{2n+1}=y_n$. Consider the continuous
function $f$, defined on
the interval $[0, +\infty),$ whose the graph passes from the points
$(\frac{1}{n}, z_n)$, it is
linear in between and vanishes at zero. It is not hard to see that $f$ is
not quiet at zero.
\par
This example may justify why the function $f$ is named "quiet". If we
discuss in details
the behavior of $f$ which is not quiet at zero, then we can observe rapid
oscillations close to
zero. We mean that as the argument approaches to zero from the right, the
rate of successive
maximum and minimum of $f$ is not bounded, though the limit might be zero.
So, $f$ has a kind of a
singular motion to the zero.

\head 3. The assumptions and some lemmas\endhead

In the sequel we shall denote by $\Bbb R$ the real line, by $\Bbb R^+$ the
interval $[0,+\infty )$
and by $I$ the interval $[0,1]$. Then $C(I)$ will denote the space of all
continuous functions
$x:I\to\Bbb R$. Let $C_{0}^{1}(I)$ be the space of all functions $x:I\to
\Bbb R$, whose the first
derivative $x'$ is absolutely continuous on $I$ and $x(0)=0$. This is a
Banach space when it is
furnished with the norm $\|\enskip \| $ defined by
$$\| x\|:=\sup\{|x'(t)|: t\in I\},\enskip x\in C_{0}^{1}(I).$$
Also we denote by $L^+_1(I)$ the space of all functions $x:I\to \Bbb R^+$
which are Lebesgue
integrable on I, endowed with its usual norm $\|\cdot\|_1$.
\par
Now consider the problem $(1.1)$, $(1.2)$, $(1.3)$. By a solution of this
problem we mean a
function $x\in C_{0}^{1}(I)$ satisfying equation $(1.1)$ for almost all
$t\in I$ and condition
$(1.3)$. \par As the functions appeared in this problem we assume the
following:
\vskip 0.2cm
\roster

\item"(H1)"  $f\colon\Bbb R\to\Bbb R$ is a continuous function, with $f(x)>
0$, when $x>0$ and
quiet at zero.

\item"(H2)" The functions $p, q$ belong to $C(I)$ and they are such that
$p>0$, $q\ge 0$ and
$\sup\{q(s): \eta \le {s}\le1\}>0$. It is clear that without loss of
generality we can assume that
$p(1)=1$.

\item"(H3)" The function $g\colon I\to \Bbb R$ is increasing and such that
$$g(\eta )=0<g(\eta +)=:b_0 .$$

\item"(H4)" It holds: $\int_{\eta}^{1}\frac{1}{p(s)}dg(s)<1.$

\endroster
\vskip .1 in
Last assumption implies that the quantity
$$\alpha :=\left (1-\int_{\eta}^{1}\frac{1}{p(s)}dg(s)\right )^{-1}$$
is a real number.
\par
As we indicated in [25], the problem
$(1.1)$, $(1.2)$, $(1.3)$ is equivalent to the operator equation  $x=Ax$,
$x\in C_{0}^{1}(I)$,
where $A$ is the operator defined by
$$Ax(t):=\alpha P(t)\int_{\eta}^{1}\Phi (f(x))(s)dg(s)+\int_{0}^{t}\Phi
(f(x))(s)ds. \tag 2.2
$$ Here the functions $P$ and $\Phi$ are defined by
$$P(t):=\int_{0}^{t}\frac{1}{p(s)}ds, \enskip t\in I$$
and
$$(\Phi y)(t):=\frac{1}{p(t)}\int_{t}^{1}q(s)y(s)ds, \enskip t\in I,
\enskip y\in C(I)$$
and the constants $H, \theta$ are given by
$$H:=\int_{\eta}^{1}\Phi (1) (s)dg(s) \enskip \hbox{and}\enskip
\theta:=\frac{p(0)}{\alpha H+\|q\|_1}.$$

It is clear that $A$ is a completely continuous operator.
\par
In the sequel we shall do use of the function
$$f_s (w):=\sup\{f(x):x\in[0,w]\},$$ for which we assume the following:
\roster

\item"(H5)"  There exists a point $v>0$ such that $f_s (v)\le \theta v.$

\endroster
\par
Now we set
$$\Bbb X:=\left\{x\in C^1_0 (I):x\ge 0, x'\ge 0,
\enskip x \enskip \hbox{is concave and}\enskip \|x\| \le v.\right\}$$
and we give the following auxiliary results:
\vskip 0.2cm
\proclaim{ Lemma 3.1}
It holds $A\Bbb X\subset \Bbb X .$
\endproclaim
\demo{Proof}
Let $x\in \Bbb X$. Then $Ax(t)\ge 0$, $(Ax)'(t)\ge 0$, and
$(Ax)''(t)=-q(t)f(x(t))\le 0$
for all $t\in I$. Also, $x\in \Bbb X$ implies that $0\le x(t)\le v$ for all
$t\in I$. Then
$$
\aligned
\|Ax\|=(Ax)'(0) &=\frac{\alpha}{p(0)}\int_{\eta}^{1}\Phi
(f(x))(s)dg(s)+\frac{1}{p(0)}\int_{0}^{1}q(s)f(x(s))ds\\
&\le f_s (v)\left [\frac{\alpha H}{p(0)}+
\frac{1}{p(0)}\int_{0}^{1}q(s)ds\right ]\\
&\le \theta v\left [\frac{\alpha H +\|q\|_1}{p(0)}\right ]= v.
\endaligned
$$
\qed\enddemo
\vskip 0.2cm
\proclaim{ Lemma 3.2}
There exists a $\lambda_v >0$ such that for all $x\in \Bbb X$ it holds
$$\int_{\eta}^{1}\Phi (f(x))(s)dg(s)\ge
\frac{b_{0}}{\lambda_v}\int_{0}^{1}q(s)f(x(s))ds.$$
\endproclaim
\demo{Proof}
From the assumption $(H3)$ we have
$$g(s)\ge b_0 , \enskip s\in (\eta,1].\tag 3.1$$
Let $x\in \Bbb X$. Then $x$ is nondecreasing and $\|x\|\le v$. Since $f$ is
quiet at zero, for the
number $T_v :=v$, there is a $\mu_v\ge 1$ such that (2.1) holds for all
$\tau \in (0,T_v )$. Hence
(2.1) also holds for the real number $\tau:=x(\eta )<\|x\| \le v$.
Therefore we have
$$
\aligned
\int_{0}^{1}q(s)f(x(s))ds
&=\int_{0}^{\eta}q(s)f(x(s))ds+\int_{\eta}^{1}q(s)f(x(s))ds\\
&\le\sup_{w\in [0,\tau
]}f(w)\int_{0}^{\eta}q(s)ds+\int_{\eta}^{1}q(s)f(x(s))ds\\
&\le\frac{\int_{0}^{\eta}q(s)ds}{\int_{\eta}^{1}q(s)ds}\enskip
\frac{\sup_{w\in [0,\tau]}f(w)}{\inf_{w\in [\tau ,T_v ]}f(w)}
\int_{\eta}^{1}q(s)f(x(s))ds\\
&+\int_{\eta}^{1}q(s)f(x(s))ds\\
&\le\left (\frac{\int_{0}^{\eta}q(s)ds}{\int_{\eta}^{1}q(s)ds}\mu_v +1\right
)\int_{\eta}^{1}q(s)f(x(s))ds\\
&=\xi \int_{\eta}^{1}q(s)f(x(s))ds,
\endaligned
$$
where $$\xi:=\frac{\int_{0}^{\eta}q(s)ds}{\int_{\eta}^{1}q(s)ds}\mu_v +1.$$
Next we use (3.1) and get
$$
\aligned
\int_{0}^{1}q(s)f(x(s))ds &\le\xi\int_{\eta}^1 q(s)f(x(s))ds\\
&\le\frac{\xi}{b_0}\int_{\eta}^1 q(s)f(x(s))g(s)ds\\
&=-\frac{\xi}{b_0}\int_{\eta}^1 d\left (\int_s^1q(r)f(x(r))dr\right ) g(s)\\
&=\frac{\xi}{b_0}\int_{\eta}^1 \int_s^1q(r)f(x(r))drdg(s)\\
&\le \frac{\lambda_v }{b_0}\int_{\eta}^1 \frac{1}{p(s)}\int_{s}^1
q(s)f(x(s))drdg(s),
\endaligned
$$
where
$$\lambda_v : =\left
(\frac{\int_{0}^{\eta}q(s)ds}{\int_{\eta}^{1}q(s)ds}\mu_v +1\right
)\sup_{s\in I}p(s).$$
The proof of the lemma is complete.
\qed\enddemo
\vskip 0.2cm
Now we set
$$D:=\int_{\eta}^{1}\Phi (P)(s)dg(s),$$
$$b:=min\left \{b_0 ,\frac{ \lambda_v H}{\alpha |D\eta p(0)-H|} \right \}$$
and
$$\sigma_v := \frac {\alpha bDp(0)}{\alpha b+\lambda_v}.$$

\proclaim{ Lemma 3.3}
It holds $$\sigma_v \eta \le H.$$
\endproclaim
\demo{ Proof}
If $D\eta p(0)-H>0$, then we have $b\le\frac{ \lambda_v H}{\alpha (D\eta
p(0)-H)}.$ Solving
with respect to $H$ we obtain the result. Also, if $D\eta p(0)-H<0$, then
$$\sigma_v \eta =\frac {\alpha bp(0)\eta} {\alpha
 b +\lambda_v}D<\frac {\alpha  bH} {\alpha  b +\lambda_v}\le H.$$
\enddemo


\head 4. Main results\endhead

 In this section we present our main results.  Let us first define the function
 $$f_i (w):=\inf \left\{f(z):\frac{\eta\sigma_v}{H}w\le z\le w\right\} ,$$
 the cone
 $$\Bbb K:=\left\{x\in C^1_0 (I):x\ge 0, x'\ge 0, \enskip x \enskip
\hbox{is concave
and}\enskip \int_{\eta}^{1}\Phi (x) (s)dg(s)\ge \sigma_v \|x\|\right\}$$
and let
$$\rho :=\frac{1}{\alpha H}.$$

\proclaim {Theorem 4.1}
Consider the functions $f,p,q$ and $g$ satisfying the assumptions
$(H1)-(H5)$ and the following
one:
\par
\quad
$(H6)$ \quad There exists $u>0$ such that $u<v$ and $f_i (u)\ge \rho u$.
\vskip 0.1cm
Then the boundary value problem $(1.1)$, $(1.2)$, $(1.3)$ admits a solution
$x$ such that
$u\le \|x\|\le v$.
\endproclaim
\demo{Proof}
Let $B_v$ be the open ball $\{x\in C^1_0 (I):\|x\|<v\}$. We claim that
$$A:\Bbb K\cap B_v\to \Bbb K .$$
Indeed, let $x\in \Bbb K\cap B_v$. Then $x\in \Bbb K$ and $x\in \Bbb X$.
First observe that $Ax\ge
0$, $(Ax)'\ge 0$ and $(Ax)''\le 0$. Moreover,
 $$
\aligned
\int_{\eta}^{1}\Phi (Ax)(s)dg(s)
&\ge \alpha \int_{\eta}^{1}\Phi (P)(s)dg(s)
\int_{\eta}^{1}\Phi (f(x))(s)dg(s)\\
&=\alpha D
\int_{\eta}^{1}\frac{1}{p(s)}\int_{s}^{1}
q(\theta)f(x(\theta ))d\theta dg(s)\\
&=\frac{\sigma_v (\alpha b_0
+\lambda_v)}{ b_0 p(0)}
\int_{\eta}^{1}\frac{1}{p(s)}\int_{s}^{1}
q(\theta )f(x(\theta ))d\theta dg(s)\\
&=\frac{\sigma_v}{p(0)}\left (\alpha +\frac{\lambda_v}
{ b_0}\right )
\int_{\eta}^{1}\frac{1}{p(s)}\int_{s}^{1}
q(\theta )f(x(\theta ))d\theta dg(s)\\
&=\sigma_v \frac{\alpha}{p(0)}\int_{\eta}^{1}
\frac{1}{p(s)}\int_{s}^{1}q(\theta)f(x(\theta ))d\theta dg(s)\\
&+\sigma_v \frac{1}{p(0)}\frac{\lambda_v}{b_0}
\int_{\eta}^{1}\frac{1}{p(s)}\int_{s}^{1}q(\theta
)f(x(\theta ))d\theta dg(s).
\endaligned$$
Hence, taking into account Lemma $3.2$ we get
$$\aligned
\int_{\eta}^{1}\Phi (Ax)(s)dg(s)
&\ge
\sigma_v \frac{\alpha}{p(0)}\int_{\eta}^{1}
\frac{1}{p(s)}\int_{s}^{1}q(\theta
)f(x(\theta ))d\theta dg(s)\\
&+\sigma_v\frac{1}{p(0)}\int_{0}^{1}q(\theta )f(x(\theta ))d\theta \\
&=\sigma_v (Ax)'(0)\\
&=\sigma_v \|(Ax)\|,
\endaligned
$$
which proves our claim.

Now consider a function $x\in\Bbb K$, with $\|x\|=u$. The fact that $x$ is
concave implies that
$$\eta x(1)\le x(\eta )\le x(r)\le x(1)\le \|x\|,
\enskip \hbox{for every}\enskip r\in [\eta , 1].$$
So,
$$
\aligned
\sigma_v\|x\| &\le \int_{\eta}^{1}\Phi (x)(s)dg(s)\\
&=\int_{\eta}^{1}\frac{1}{p(s)}\int_{s}^{1}q(\theta
)x(\theta )d\theta dg(s)\\
&\le x(1)\int_{\eta}^{1}\frac{1}{p(s)}
\int_{s}^{1}q(\theta )d\theta dg(s)\\
&=x(1)\int_{\eta}^{1}\Phi (1)(s)dg(s)\\
&=x(1)H.
\endaligned
$$
Thus we have
$$x(1)\ge \frac{\sigma_v\|x\|}{H},$$
which implies that
$$x(r)\ge \frac{\eta \sigma_v}{H}\|x\|, \enskip r\in [\eta,1].$$
Therefore, for every $r\in [\eta, 1]$ we have
$$\frac{\eta \sigma_v}{H}\|x\|\le x(r)\le \|x\| ,$$
where, notice that it also holds $\frac{\eta \sigma _v}{H}\le 1,$ see Lemma
3.3.  Then, by
assumption $(H6)$, we obtain
$$
\aligned
(Ax)'(1)&\ge \alpha\int_{\eta}^{1}\frac{1}
{p(s)}\int_{s}^{1}q(\theta)f(x(\theta ))d\theta dg(s)\\
&\ge \alpha f_i (u) H \\
&\ge \alpha H\rho u =u .
\endaligned
$$
This means that, if $\|x\| =u$, then $\|Ax\|\ge \|x\| .$ Moreover in Lemma
3.1 we have proved that if
$\|x\| =v$, then $\|Ax\|\le \|x\| .$\par To complete the proof we set
$\Omega _1 :=\{x\in C_{0}^{1}(I): \|x\|<u\}$, $\Omega _2 :=B_v$ and apply
Theorem 1.1.

\qed\enddemo
An immediate consequence of this theorem is the following:\vskip 0.2cm
\proclaim {Corollary 4.2}
Consider the functions $f, q, g$ satisfying the assumptions  $(H1)-(H4)$.
Moreover assume that

\roster

\item"(H7)"  There exist two two-sided sequences $(u_ {_k})$, $(v_ {_k})$,
$k\in \Bbb Z$ such that
$$0< u_ {_k} <v_{_k} <u_ {_{k+1}},$$
$$f_i (u_ {_k} )\ge \rho u_ {_k}\enskip \hbox{and}\enskip f_s (v_ {_k} )\ge
\theta v_ {_k}$$
for every $k\in \Bbb Z$.
\endroster

\par
Then there exists a sequence $x_{_k}$, $k\in \Bbb Z$ of solutions of the
boundary value
problem $(1.1)$, $(1.2)$, $(1.3)$, such that
$$u_ {_ k} <\|x_ {_ k}\| <v_ {_k}, \enskip k\in \Bbb Z.$$

\endproclaim

\Refs
\refstyle{A}
\widestnumber\key{33}

\ref\key 1
\by  R. P. Agarwal and D. O'Regan
\paper Twin solutions to singular Dirichlet problems
\jour J. Math. Anal. Appl. \vol 240 \yr1999
\pages 433--445
\endref


\ref\key 2
\by  R. P. Agarwal and D. O'Regan
\paper Twin solutions to singular boundary value problems
\jour Proc. A.M.S. \vol 128 \yr2000
\pages 2085--2094
\endref

\ref\key 3
\by  V. Anuradha, D. D. Hai and R. Shivaji
\paper Existence results for superlinear semipositone BVP'S
\jour Proc. A.M.S. \vol 124 \yr1996
\pages 757--763
\endref

\ref\key 4
\by  R. Avery
\paper Existence of multiple positive solutions to a conjugate boundary
value problem
\jour Math. Sci. Rcs. Hot-Line \vol 2 \yr1998
\pages 1--6
\endref

\ref\key 5
\by  R. Avery
\paper A generalization of the Legget-Williams fixed point theorem
\jour Math. Sci. Rcs. Hot-Line \vol 3 \yr1999
\pages 9--14
\endref

\ref\key 6
\by  R. Avery, J. M. Davis and J. Henderson
\paper Three symmetric positive solutions for Lidstone problems by a
generalization of
the Legget-Williams theorem
\jour Electron. J. Differential Equations \vol 2000, No. 40 \yr2000
\pages 1--21
\endref

\ref\key 7
\by  R. Avery and J. Henderson
\paper Three symmetric positive solutions for a second order  boundary
value problem
\jour Appl. Math. Lett. \vol 13 \yr2000
\pages 1--7
\endref

\ref\key 8
\by  A. V. Bitsadze
\paper On the theory of nonlocal boundary value problems
\jour Soviet Math. Dock. \vol 30 \yr1964
\pages 8--10
\endref

\ref\key 9
\by  A. V. Bitsadze and A. A. Samarskii
\paper Some elementary generalizations of linear elliptic boundary value
problems
\jour Dokl. Akad. Nauk SSSR \vol 185 \yr1969
\pages 739--740
\endref

\ref\key 10
\by G. Bonanno
\paper Existence of three solutions for a two point boundary value problem
\jour Appl. Math. Lett.  \vol 13 \yr2000
\pages 53--57
\endref

\ref\key 11
\by   C.J. Chyan and J. Henderson
\paper Multiple solutions for $2mth$ Sturm-Liouville boundary value problem
\jour Comput. Math. Appl. \vol 40 \yr2000
\pages 231--237
\endref

\ref\key 12
\by J. M. Davis, P.W. Eloe and J. Henderson
\paper Triple positive solutions and dependence on higher order derivatives
\jour J. Math. Anal. Appl. \vol 237 \yr1999
\pages 710--720
\endref

\ref\key 13
\by  K. Deimling
\book Nonlinear Functional Analysis
\publ   Springer-Verlag, New York \yr 1985
\endref



\ref\key 14
\by  D. R. Dunninger and H. Wang
\paper Multiplicity of positive solutions for a nonlinear differential
equation with
nonlinear boundary conditions
\jour Annales Polonici Math. \vol LXIX.2 \yr1998
\pages 155--165
\endref

\ref\key 15
\by  P. W. Eloe and J. Henderson
\paper Positive solutions and nonlinear multipoin conjugate eigenvalue
problems
\jour Electronic J. Diff. Equations  \vol 1997, No. 8 \yr1997
\pages 1--11
\endref

\ref\key 16
\by  L. H. Erbe and H. Wang
\paper On the existence of positive solutions of ordinary differential
equations
\jour Proc. A.M.S. \vol 120 \yr1994
\pages 743--748
\endref

\ref\key 17
\by  L. H. Erbe, S. Hu  and H. Wang
\paper Multiple positive solutions of some boundary value problems
\jour J. Math. Anal. Appl. \vol 184 \yr1994
\pages 640--648
\endref

\ref\key 18
\by  D. Guo and V. Lakshmikantham
\book Nonlinear Problems in Abstract Cones,
\publ  Academic Press, Orlando, Fl. \yr 1988
\endref

\ref\key 19
\by  C. Gupta
\paper Solvability of a generalized multi-point boundary value problem of
mixed type for second
order ordinary differential equations
\jour Proc. Dynamic Dystems and Applications \vol 2 \yr1995
\pages 215--222
\endref

\ref\key 20
\by  C. Gupta, S. K. Ntouyas and P. Ch. Tsamatos
\paper On an m-point boundary value problem for second order ordinary
differential equations
\jour Nonlinear Analysis TMA \vol 23 \yr1994
\pages 1427--1436
\endref

\ref\key 21
\by  J. Henderson and H.B. Thompson
\paper Multiple symmetric positive solutions for a second order  boundary
value problem
\jour Proc. A.M.S. \vol 128 \yr2000
\pages 2373--2379
\endref

\ref\key 22
\by  J. Henderson and E. R. Kaufmann
\paper Multiple positive solutions for focal  boundary value problems
\jour Comm. Appl. Anal. \vol 1 \yr1997
\pages 53--60
\endref

\ref\key 23
\by  V. Il'in and E. Moiseev
\paper Nonlocal  boundary value problems of the second kind for a
Sturm-Liouville operator
\jour Differential Equations \vol 23 \yr1987
\pages 979--987
\endref

\ref\key 24
\by  D. Jiang
\paper Multiple positive solutions to singular boundary value problems for
superlinear
higher-order ODEs
\jour Comput. Math. Appl. \vol 40 \yr2000
\pages 249--259
\endref

\ref\key 25
\by  G. L. Karakostas and P. Ch. Tsamatos
\paper Positive solutions for a nonlocal boundary value problem with
increasing response
\jour Electron. J. Differential Equations
\vol 2000, No. 73
\yr 2000
\pages1-8
\endref

\ref\key 26
\by  G. L. Karakostas and P. Ch. Tsamatos
\paper On a nonlocal boundary value problem at resonance
\jour J. Math. Anal. Appl. (in press)
\pages
\endref

\ref\key 27
\by  G. L. Karakostas and P. Ch. Tsamatos
\paper Existence results for some n-dimensional nonlocal boundary value
problems
\jour J. Math. Anal. Appl. (in press)
\pages
\endref

\ref\key 28
\by  G. L. Karakostas and P. Ch. Tsamatos
\paper Positive solutions of a boundary-value problem for second order
ordinary differential
equations
\jour Electron. J. Differential Equations
\vol 2000, No. 49
\yr 2000
\pages 1-9
\endref

\ref\key 29
\by  L. Kong and D. Jiang
\paper Multiple positive solutions of a nonlinear fourth order periodic
boundary value problem
\jour Annales Polonici Math.
\vol LXIX.3
\yr1998
\pages 265--270
\endref

\ref\key 30
\by  M. A. Krasnoselskii
\book Positive solutions of operator equations
\publ  Noordhoff, Groningen
\yr 1964
\endref

\ref\key 31
\by R. W. Legget and L.R. Williams
\paper A fixed point theorem with application to an Infectius disease model
\jour J. Math. Anal. Appl. \vol 76 \yr1980
\pages 91--97
\endref

 \ref\key 32
\by  R. Ma
\paper Multiplicity of positive solutions for second-order three-point
boundary value
problems
\jour Comput. Math. Appl. \vol 40 \yr2000
\pages 193--204
\endref

\ref\key 33
\by B. Ricceri
\paper On a three critical points theorem
\jour Arch. Math.(Basel)  \vol 75  \yr2000
\pages 220--226
\endref

\endRefs


\enddocument
